https://scholar.harvard.edu/files/dlcammack/files/dnt5_1.26.17.

pdf

NOT TALKING BUT THINKING AND VOTING:

DEMOCRATIC DELIBERATION IN CLASSICAL ATHENS

Daniela Cammack1,2

Abstract. Classical Athenian democracy is often described as deliberative, implying that discus-
sion by the dmos played an important political role. But of the three Greek verbs associated with
deliberation, only one, bouleuomai, denoted an action performed by the dmos, and in mass politi-
cal contexts it suggested not discussion but internal decision-making communicated by voting.
While speech was crucial to democratic politics, it was oratorical rather than dialogical and per-
formed by rhtores, orators or politicians, who by the very act of speaking were conceived as
casting themselves outside the deliberating dmos. With respect to public speech, classical Atheni-
an democracy had more in common with modern democratic politics than is usually recognized.
This similarity makes it more, not less, useful as a model today.

Classical Athenian democracy is often described as deliberative, implying that

discussion by the dmos played an important political role.3 This is not a neces-
sary implication. As Robert Goodin and others have emphasized, deliberation
may also suggest internal thought, as it does in Aristotle, Hobbes and Rousseau.4
Since the nineteenth century, however, political deliberation has typically been
associated with speech. John Stuart Mill defined the function of a deliberating
body as to secure hearing and consideration to many conflicting opinions, while
John Rawls invoked the internal type but shifted to dialogue (common or pub-
lic deliberation) when discussing legislation.5 Most influential has been Jrgen
Habermas, whose work on communicative action effectively launched a new sub-
field of political theory.6 Those following his lead have typically defined delibera-
tion as discussion, often but not always undertaken with a view to making a deci-
sion or recommendation on a public matter, where the give-and-take of argument
among diverse participants will, it is hoped, lead to a better understanding of the
issues involved and thus to a more reasonable outcome.7 Such discussion has been
deemed essential for democratic legitimacy.8

1
From July 2017, Assistant Professor, Political Science, Yale University. This draft Jan. 26,
2017. Email: daniela.cammack@yale.edu.
2
I thank Daniel Betti, Paul Cammack, Stefan Eich, Conor Farrington, Bryan Garsten, Robert
Goodin, David Grewal, Beth Janairo, Melissa Lane, David Langslow, Jane Mansbridge, Josiah
Ober, Gary Remer, David Runciman, Daniel Schillinger, George Scialabba, Peter Spiegler, Paul
Steedman, Richard Tuck, Philippe Urfalino, two anonymous referees and audiences at Harvard,
Cambridge, and Yale. I have used the Loeb Library for Greek texts and often, to illustrate transla-
tion practices, its English renderings as well. Unattributed translations are my own.
3
Balot 2004, pp. 63-6; Elster 1998, pp. 1-2; Fishkin 2009, pp. 11-13; Gutmann and Thomp-
son 2004, p. 8; Ober 2008, pp. 161-7, 264, 279. Cf. Arendt 1998 [1958], p. 26; Finley 1985, p. 13.
4
Aristot. Eud. Eth. 1227a18, Nic. Eth. 1112b15; Hobbes Civ. 13.16, Lev. I.6; Rousseau Cont.
Soc. II.3.3. Goodin 2003, 2008; Manin 2011; Stevens 1933.
5
Mill 1991 [1861], p. 272; Rawls 1999, pp. 138-9, 283, 315-16.
6
Habermas 1985, 1989, 1996.
7
Benhabib ed. 1996; Bohman 1996; Dryzek 2000; Fishkin and Laslett 2003; Gutmann and
Thompson 1996, 2004; Manin 1987; Mansbridge et al. 2012. Bchtiger et al. 2010 examine the
diversity of conceptions of deliberative democracy. On the relationship between deliberation and
deciding, compare Habermas 1985, 1996; Gutmann and Thompson 1996, 2004; Urfalino 2010.
8
Benhabib 1996; Cohen 1989; Manin 1987; Lafont 2015.

Electronic copy available at: https://ssrn.com/abstract=2161074

 Athens is regarded as a precedent for several reasons. As is well known, most
citizens had the right to speak publicly (isgoria),9 and they could say more or
less what they pleased (parrhsia).10 Formal political debates could be long and
energetic, and written speeches form a large part of the surviving literary evi-
dence.11 Two items are particularly often cited: Aristotles defence of rule by a
multitude, which has been interpreted as concerned with diverse speech, and a
remark by Pericles to the effect that Athenians saw speech (logos) not as a stum-
bling-block in the way of action but rather as an indispensable preliminary to it.12
Most compelling, or so it would seem, the Athenian dmos is found deliberat-
ing, discussing and debating throughout our translated texts,13 while refer-
ences to deliberative speakers and deliberative rhetoric are also common.14
Combined with evidence of informal discussion about politicsas suggested, for
example, by Platos dialoguesone may wonder if Athenian citizens did much
other than engage in political speech.
Yet the evidence for dialogue as opposed to oratory in the Athenian political
process is surprisingly weak. Much of the above evidence could as easily apply to
the kind of plebiscitary rhetoric characteristic of the Roman Republic as to
group discussion, and contextual details support the former.15 The assembly typi-
cally included some five to eight thousand citizens, rising to fourteen or even
twenty thousand by the end of the democratic period.16 Genuine group discussion
was thus impossible. Virtually all assemblygoers attended to listen and vote, not
to speak.17 Knowing this, several scholars have identified the council of five hun-
dred as Athens principal deliberative venue, but intriguingly, as we shall see, its
activity was seldom described as deliberation.18 Athens democratic courts raise
another question. Judges were often asked to deliberate well (bouleuesthe

9
Hdt 5.78; Eur. Supp. 438, 441; Ps. Xen. Ath. Pol. 1.2; Plat. Gorg. 461e, Prot. 319b-d, 322d-
23a; Dem. 15.18, 20.105-6; Aeschin. 3.20; Finley 1985, pp. 18-19; Lewis 1971; Miller 2001;
Monoson 2000, pp. 56-60; Ober 1989, pp. 72-9, 295-7; Sluiter and Rosen ed. 2004.
10
Eur. Ion 671-5, Hipp. 420-3, Supp. 433-41; Plat. Rep. 557b, Gorg. 461c; Dem. 9.3, 60.25-6;
Ps. Xen. 1.2; Ps. Aristot. 16.6; Isoc. 2.3-4, L4.6. Landauer 2012; Miller 2001; Monoson 2000, pp.
52-63; Radin 1927; Saxonhouse 2006, pp. 85-128; Sluiter and Rosen eds. 2004.
11
Thuc. 3.36-50; Xen. Hell. 1.7. Ober 1989, pp. 341-9, catalogues 104 speeches addressed to
courts, 27 to assemblies and 5 to councils, 5 mass funeral orations, and 22 related texts.
12
Aristot. Pol. 1281a-b, 1286a25-32; Thuc. 2.40.2 (cf. Lys. 2.19; Xen. Hell. 1.1.23). Dahl
1989, p. 18; Dunn 2005, pp. 27, 178; Elster ed. 1998, p. 1; Gutmann and Thompson 1996, p. 131,
2004, p. 8; Held 2006, pp. 13-15; Urbinati 2002, epigraph; Waldron 1999, pp. 92-123.
13
Aeschin 2.109-10, 60, 134, 3.71, trans. Adams; Dem. 10.10, 15.1, 31, 18.73, 165, 25.9,
trans. Vince; Isoc. 13.2, trans. Norlin; Thphr. Char. 26.1, trans. Rusten.
14
Aristot. Rh. 1354b, 1358b-62a, 1369b, 1377b, 1413b, 1415b trans. Barnes, Freese, Lawson-
Tancred; Lys. 14.45 trans. Lamb; Dem. 19.15, 22.48, 23.13 trans. Vince; Aeschin. 1.1, 195, 2.56
trans. Adams; Din. 2.15, 17 trans. Burtt.
15
The phrase is Chambers (2009). This point is emphasized by Manin 1997, 2011; I develop
his arguments by establishing the distinction between speakers, who advised, and audience, which
deliberated. Cf. Finley 1985, p. 56; Remer 1999, 2000; Urfalino 2005.
16
Forsn and Stanton eds. 1996; Hansen 1999, 130-2.
17
As emphasized by Elster 1998, p. 2; Fishkin 2009, p. 12; Finley 1985, p. 24; Hansen 1989,
pp. 93-127, 1999, pp. 142-7; Johnstone 1997, p. 110; Ober 1989, pp. 324-5, 2008, p. 164; Rhodes
1972, p. 80. Cf. Dahl 1989, 225-31.
18
Fishkin 2009, pp. 11, 82; Ober 1989, pp. 142-51; Ober 2013; Ruz 1997, pp. 437-543.

Electronic copy available at: https://ssrn.com/abstract=2161074

kals), but this did not imply speech.19 Aristotle reports that most Hellenic legisla-
tors did not let judges speak together (koinlogntai) and Athens was no excep-
tion. Once both sides of a case had been heard, the judges simply collected their
ballots and voted secretly.20 Nothing prevented them from making noise during
speeches (often referred to as thorubos, uproar) or from exchanging comments
on their way to the voting urns, but the deliberations to which litigants referred
were largely internal.21 Might this have been the case more generally?
This investigation requires some care with terminology. Three Greek verbs
are associated with the translation deliberative. One, dmgore, speak before
the people, address the assembly, orate or more negatively harangue, cer-
tainly indicated public speech, although not discussion.22 Deliberate is thus nev-
er used to translate this verb. But deliberative is found in translations of related
nouns and adjectives such as h dmgoria, deliberative oratory, ho dmgoros,
deliberative orator, and h rhtorik dmgorik, deliberative rhetoric.23
Symbouleu, advise or counsel, is treated similarly. As with dmgore
there is no doubt that it indicated communication, though again the action was
one-way.24 Accordingly, the verb is not translated deliberate. But deliberative
is often used to translate ho symbouleun, advisor, counsellor or deliberative
speaker, and h rhtorik symbouleutik, deliberative rhetoric.25
The only Greek verb conventionally translated deliberate is bouleuomai,
while to bouleuesthai and h bouleusis lie behind most appearances of delibera-
tion and bouleutikos behind many of deliberative. Most important, of the three
verbs just mentioned, only bouleuomai denoted an action performed by the
dmos. Dmgore and symbouleu were associated with individuals, but the sub-
ject of bouleuomai could be an individual, a small group, or a mass.26 Still more
interesting, the actions represented by dmgore or symbouleu and bouleuomai
were frequently performed by solo speakers and mass audiences simultaneously.27

19
Antiph. 5.71-3, 90-4; Lyc. 1.2, 11, 14-15, 83; Aeschin. 3.255; Din. 1.26, 98. Cf. Andoc. 4.7;
Lys. 6.8, 9.15, 19, 21.13, 25.21, 23, 28.16; Isoc. 5.170, 178, 248; Dem. 20.15, 35, 25.14, 48.52,
53.29; Aristot. Pol. 1286a25-30, 1287b30-5; Hyp. 5, fr. 3.
20
Aristot. Pol. 1268b5-10. Boegehold 1995, pp. 21-42.
21
Andoc. 1.69-70; Isoc. 15.272; Dem. 18.52, 23.19, 47.44, 50.3, Ex. 5, 21, 26; Lyc. 1.52, 58,
127; Hyp. 1.2, 20, 4.41; Thphr. Char. 7; Bers 1985; Schwartzberg 2010; Boegehold 1995, p. 39.
22
From dmos, people, and agoreu, speak in the agora. Aeschin. 1.3, 3.71, trans. Adams;
Dem. 21.202, trans. Vince; Isoc. 8.75, trans. Norlin; Dem. Ex. 9, trans. DeWitt; Plat. Gorg. 519d,
trans. Lamb. Cf. Aeschin. 1.1, 2, 3, 19-20 et passim, 3.220; Ps-Aristot. Ath. Pol. 28.3; Isoc. L1.9,
L8.7, Panath. 9; Plat. Gorg. 482c, 519d.
23
Aristot. Rh. 1354b, 1413b, trans. Lawson-Tancred; Lys. 14.45 trans. Lamb; Dem. 19.15,
22.48, 23.13 trans. Vince; Aeschin. 1.1, 195, 2.56 trans. Adams; Din. 2.15, 17 trans. Burtt.
24
Isoc. L1.2, L2.1, L6.6; Dem. Ex. 3; Din. 1.35; Aeschin. 1.110-11.
25
Aristot. Rh. 1358b-62a, 1369b, 1377b, trans. Barnes, Kennedy, Lawson-Tancred. Yunis
1996, p. 16, n. 24, discusses the shift in translating symbouleutikon from political to delibera-
tive [rhetoric]. I suggest advisory for symbouleutikon and public for dmgorikon.
26
Individuals: Hom. Il. 2.114; Aristot. NE 1112b12-18; Isoc. L2.20. Small groups: Hdt. 3.71-
84; Plat. Crit. 119d, 120c-d, Gorg. 487c-d; Dem. 19.122. Masses: Aeschin. 1.22, 33, 2.50, 60-1,
67, 70, 82, 120, 3.67, 151, 251, 255; Dem. 3.1, 5.2, 6.28, 9.19, 10.30, 18.86, 196, 19.96, 23.109,
24.99, 25.14, Ex. 6; Plat. Prot. 319a-d.
27
Plat. Prot. 319b-d, 322d-24c, Gorg. 455b-56a, Alc. I 106c-7d; Thuc. 3.43, 38; Xen. Mem.
3.6; Hell. 1.7.16, 19, 2.2.15, 2.4.40; Andoc. 4.12; Lys. 6.54, 25.27, 33.3; Isoc. 3.19, 21, 4.3, 19,

3
 Any investigation of democratic deliberation in ancient Greece must centre on
the meaning of bouleuomai. The use of deliberative to render cognates of
dmgore and symbouleu is misleading, since those verbs did not suggest dis-
cussion. To be sure, it is not strictly necessary to infer from these translations that
the deliberation implied is dialogical rather than internal. Deliberative rhetoric,
for example, may suggest rhetoric intended to assist the dmos in its internal de-
liberations. Indeed, when such translations were first advanced, in the nineteenth
century and earlier, that was possibly what the translators had in mind. Even so,
these renderings seem ill-advised. That is because when deliberate, delibera-
tive and deliberation appear in connection with both speakers and listeners, the
impression naturally arises that speakers and listeners are engaged in the same
activity. On the dialogical interpretation of deliberation, that looks fine: both
speakers and listeners are indeed part of the deliberating group. But in our ancient
Greek texts, typically, speakers alone orate (dmgore) and advise (symbouleu),
while the audience alone deliberates (bouleuomai).
This article argues that in mass political contexts in classical Athens, delibera-
tion implied internal thought rather than interpersonal talk. I begin by using large-
ly non-Athenian evidence to establish three models of deliberation represented by
bouleuomai, one internal, one dialogical and one mixed. The arguments presented
are both philosophical and philological; the latter material may be unfamiliar to
some readers, but I hope not alienating. Next, I analyze Athens courts, council
and assembly in the light of this framework and other historical and textual evi-
dence, and find that dialogical deliberation, at least on the ancient Greek concep-
tion, did not take place in any of them.28
Athenian democracy should thus not be treated as a historical precedent for
deliberative democracy as it is conceived today. Yet the study of Athenian delib-
eration still has value for political theorists. For one thing, it offers a useful per-
spective on the relationship between deliberation and decision-making.
Bouleuomai specifically denoted coming to a decision about a course of action
within the deliberators power, and unlike some modern uses of deliberate, it
always referred to both the consideration stage and the moment of decision. To
consider an action without deciding itor to consider issues not open to being
altered by human actionwas not, in ancient Greek, to deliberate at all, but to
contemplate (there) or to converse (dialog). The implication for democracy
was clear. If large numbers of people were to deliberate politically, i.e. to make
political decisions, the deliberation had to be internal and culminate in a vote.
In addition, the fact that most Athenians participated in politics by voting ra-
ther than speaking suggests a greater similarity between classical Athenian and
modern democratic politics than is often recognized. Athens was too large for

170-1, 5.88, 143, 7.77-8, 8.1-2, 8-13, 27, 52-55, 12.170-1, 15.256; Aeschin. 1.26, 64, 1.110-11,
120, 180, 186, 2.29, 49, 65, 157, 165, 3.158, 225-6; Dem. 4.1, 8.1, 73, 10.17, 75, 14.8, 15.1, 18.86,
Ex. 1, 3, 6, 11, 20, 26-7, 30, 33, 35-6, 56; Din. 1.31, 35-6, 40, 72, 76-8, 81, 93, 2.14, 15; Hyp. 5
col. 28; Ps. Aristot. Ath. Pol. 23-4, 29.
28
Another venue worth examining would be Athens local assemblies or dmoi: cf. Ober
1989, p. 106; Osborne 1985; Whitehead 1986. What little evidence we have supports my case: e.g.
the use of dmgore, not bouleuomai, to describe speech-making at Dem. 57.9.

4
genuine public conversation.29 Many political tasks, including speech-making,
were consequently outsourced to small numbers of individuals eager to increase
their personal honour and renown.30 The most important differences between
classical Athenian and modern democracy are thus not the use of dialogical delib-
eration in the political process but the fact that in Athens, all political decisions
were made by very large groups of ordinary citizens, under advice from those
who sought a leading political role, and that the bar to performing a leading role
was low, while the risks associated with it were highthe opposite of the case
today. This system would not be beyond our power to imitate, if we chose.

I. Three Models of Deliberation

Bouleuomai is widely attested from Homer to Aristotle. We shall begin, therefore,

by exploring its use outside Athenian political settings. Three models of delibera-
tion emerge. The first, both chronologically and as presented philosophically by
Aristotle, was internal deliberation by a single agent. Dialogical deliberation, fea-
turing small group discussion, also appears from the fifth century on. A third type,
which we may call guided deliberation, featured an internally deliberating deci-
sion-maker who took advice from others. Though seldom discussed, the last is the
most important for understanding the Athenian democratic process, and may
equally have the most to offer political theorists today.

Introducing bouleuomai: Aristotles account

Aristotle provides the only surviving philosophical examination of h bouleusis or
to bouleuesthai, deliberation.31 His account is clear, consistent, supported by a
noteworthy feature of Greek grammar and uncontradicted by any other evidence,
though it does need to be supplemented. Deliberation, as represented by Aristotle,
was typically internal; it meant coming to a decision about a course of action
within the deliberators power; it was a two-stage process, involving first consid-
ering, then deciding; and while it could be performed by groups, Aristotle gave no
sign that such cases had to or even might include discussion.
That Aristotle took the paradigm of deliberation to be internal is shown by the
fact that all his exemplars are single mena doctor, a general, a gymnastic train-
er, a Lacedaimonian, an orator, a prudent man, Pericles.32 Though the first person
plural (bouleuometha, we deliberate) also appeared in his account, by we Aris-
totle usually meant people in general rather than a group acting together, and that
was the case here. There were exceptions: in the Nicomachean Ethics he said that
all particular divisions of men deliberate (bouleuontai) about things attainable by
their own actions, while in the Eudemian Ethics he observed that we do not de-
liberate (bouleuometha) on affairs in India, presumably thinking of Greek com-

29
Pace Gutman and Thompson 1996, p. 131.
30
See Finley 1985, pp. 38-75.
31
Significant discussions include Bickford 1996, pp. 25-54; Hardie 1980; Irwin 1975; Mulgan
1999; Yack 1996. See also Cammack 2013b.
32
Aristot. Eud. Eth. 1227a19, 1227b25; Nic. Eth. 1112a32, 1112b12-14, 1140a26, 1140b6.
Cf. Hist. Anim. 488b25; Pol. 1260a1.

5
munities rather than single citizens.33 In the Rhetoric and Politics, too, the gram-
matical subject of bouleuomai was often a group.34 But nothing in Aristotles
philosophical analysis required the presence of others. Its focus was the delibera-
tor (ho bouleuomenos) in the singular.35
What did this activity involve? According to Aristotle, we deliberate exclu-
sively about things from us and from action, or more elegantly practical matters
within our power.36 We do not deliberate about eternal things such as geomet-
rical truths, regular things such as solstices, irregular things such as the weather,
random things such as finding treasure, or anything else caused by nature, neces-
sity or chance.37 We deliberate only about outcomes attainable by human agency
and not even most of those. A Lacedaemonian would not deliberate about the best
political system for the Scythians, for Scythian government is not under his con-
trol.38 Equally, we do not deliberate about spellings: we may wonder how to spell
a word correctly, but the correct spelling is not up to any one of us.39 Deliberation
concerns solely things within the power of the deliberating agent to effect, via ei-
ther his own agency or that of others under his direction.40 Its subject matter is
limited because its purpose is limited. Deliberation decides the action of the de-
liberator. It is thus inextricably linked to agency and choice. It presupposes that
the deliberator is choosing between at least two possible courses of action.41
This may seem a surprisingly narrow account. Aristotle would have been puz-
zled by the idea of deliberating in a seminar room: it would have struck him as a
misuse of terms.42 On his understanding, philosophical, historical and scientific
questions cannot be deliberated about, because they cannot be influenced by the
actions of the deliberator. That is not to say that such questions cannot be consid-
ered, examined, contemplated, discussed and so on. Ancient Greek, like English,
recognized many kinds of intellectual activity. Among the terms Aristotle used to
describe his characteristic occupation were thero, theorize or contemplate,
skeptomai, consider, and skopeu, examine.43 But bouleuomai he reserved for
practical decision-making, excluding even the process of settling on an opinion.44
Though some form of imagination is common to all animals, he said in On the
Soul, the deliberative (bouleutik) imagination belongs only to those that decide
whether to do this or that.45

33
Aristot. Nic. Eth. 1112b37; Eud. Eth. 1226a28.
34
Aristot. Pol. 1275b20-25, 1281b25-30, 1286a27, 1298a4-35, 1298b10-35, 1322b37; Rh.
1358b.
35
Aristot. Eud. Eth. 1227a6-8; Nic. Eth. 1142a34.
36
Aristot. Nic. Eth. 1112a30-1. Cf. Eud. Eth. 1226a26-33; Rh. 1357a.
37
Aristot. Nic. Eth. 1112a20-30.
38
Aristot. Nic. Eth. 1112a36.
39
Aristot. Nic. Eth. 1112b1-4.
40
Aristot. Nic. Eth. 1112b28.
41
Aristot. Eud. Eth. 1227a18-19, NE 1113a2-12. See further Cammack 2013b.
42
E.g. Bohman 1996, p. 3.
43
There: Hist. Anim. 491a, Met. 1001a4, Nic. Eth. 1104a12, 1139a1, Pol. 1334b5; skepto-
mai, Aristot. Eud. Eth. 1214a15, 1217b1, 1218b27, Pol. 1324a14; skopeu: Eud. Eth. 1214b29,
1217b16, Nic. Eth. 1112a10.
44
Aristot. Nic. Eth. 1111b30-12a13.
45
Aristot. De An. 434a7-9 (italics mine).

6
 Deliberation, according to Aristotle, thus involved coming to a decision about
an action within ones power. And this, importantly, was a two-stage process. In
English, coming to a decision is ambiguous: either the consideration performed
prior to the decision or the final act of decision-making may be meant. There was
less ambiguity about bouleuomai, because the stage reached was often indicated
by tense.46 In the present tense, bouleuomai could suggest either the entire delib-
erative process, both considering and deciding, or more specifically the process of
consideration alone; in past tenses, it specified decide after consideration.47
The present participle, for instance, was used when the deliberator was still
making up his mind. The deliberator (ho bouleuomenos, present) always deliber-
ates for the sake of somethinghe always has some aim in view.48 The aorist (or
another past tense) confirmed that the decision had been made. The weak, having
come to a resolution (bouleusamenoi, aorist), on account of passion do not keep to
what they decided (ebouleusanto, aorist). The impetuous, on the other hand, on
account of not making a resolution (to m bouleusasthai, aorist), are led by pas-
sion throughout.49 Neither the weak nor the impetuous, on this account, fail en-
tirely to deliberate: the weak actually complete both parts of the deliberative pro-
cess and form a decision, though they fail to execute it, while the impetuous may
begin but do not complete the decision-making process. And the stage reached is
shown by tense.
This leads to a significant point. The fact that bouleuomai represents a two-
stage process may hardly matter when the deliberator is a single person, since
both considering and deciding are internal and the transition between them may
be virtually seamless. When the deliberating agent is a group, however, it matters
quite a lot, because different processes will necessarily feature at each stage.
Considering may involve internal thought, hearing speeches, or group discus-
sion; deciding may involve voting, either publicly or in secret, or establishing a
verbal consensus. Any combination of these options could be represented by
bouleuomai. The modern English deliberate is different in this respect, since
even in past tenses it may refer to the consideration stage alone, with a different
term (e.g. deciding) used to specify the moment of decision.50
This poses a problem for those interested in the mechanics of mass delibera-
tion in ancient Greece. Bouleuomai tells us only that a decision is reached; if we
want to know how, we need more information. What can we glean from Aristotle?
Only that nowhere in his dedicated philosophical analysis nor anywhere else does
he associate bouleuomai with group discussion.
Groups certainly did deliberate. Writing of court cases, Aristotle confirms that
panels of citizens (syniontes, those gathered together) hear trials (dikazousin),

46
Including in parts of speech that do not show tense in English, such as participles and in-
finitives. In such cases tense shows not time but aspect: progressive/repeated aspect by the present
tense, simple aspect by the aorist, and completed aspect by the perfect.
47
Liddell et al. 1996, s.v. bouleu.
48
Aristot. Eud. Eth. 1227a6-7.
49
Aristot. Nic. Eth. 1150b18.
50
As illustrated by e.g. Yack 1996, p. 420. Sintomer 2010 notes that the association of delib-
eration with decision-making remains strong in modern Italian and Portuguese. Conversely, the
German deliberative Stimme (deliberative voice) is strictly consultative (p. 485, n. 16).

7
deliberate (bouleuontai), and judge (krinousin).51 Yet, as already noted, according
to Aristotle judges did not confer.52 Groups also made policy. Aristotle observes
that decisions (kriseis) about war and peace, alliances, laws and so on may be as-
signed to all citizens, to some, or to single officials, and he adds that allowing all
to deliberate (bouleuesthai) about everything was standard in democracies of his
day.53 But he never implies that such deliberation involved group discussion.54
The defence of rule by a multitude in Politics III has been interpreted as
though it did. Aristotle argues that many men can contribute more of a certain po-
litically salient thing than a small number or a single man, and this contribution
has been interpreted as diverse speech.55 But as I have argued elsewhere, that
reading must be mistaken.56 The specific tasks that Aristotle assigns to the multi-
tude are deciding elections, audits (the routine inspection of office-holders after
their tenure) and trials, and none of these involved speech on the part of the deci-
sion-makers.57 Elections were typically held in assemblies but without debate,
while audits (usually) and trials were decided by judges, who as we have just ob-
served did not engage in discussion.58 Most important, Aristotles remark that his
argument would also apply to animals proves that speech played no part in it.59
Logos, verbal reason, was to Aristotle the essential difference between human be-
ings and other animals.60 Consequently, it cannot have been the politically salient
thing he had in mind. Aret, virtue or excellence, aggregated and even ampli-
fied through collective action, seems a more plausible candidate.61
Another line, from Politics IV, may also be interpreted dialogically: They
will deliberate (bouleusontai) better when all deliberate (bouleuomenoi) together,
the dmos with the notables and they with the majority.62 The present participle
in all deliberate together suggests that Aristotle was thinking of the considera-
tion stage, but what is meant by together is unclear. Group discussion is possi-
ble, but so is internal deliberation guided by oratory, as in court cases. Moreover,
the rest of the passage suggests a different interpretation. Aristotle was consider-
ing how to improve to bouleuesthai, here best translated policy-making, in
communities where the dmos was maximally powerful, and to this end he listed
several possible measures: increasing the number of elite citizens on the delibera-
tive body; fining elite citizens for non-attendancethe immediate prompt for the
line above; electing deliberators or choosing them equally by lot from each class;
giving pay for attendance to no more than the number of non-elite citizens needed

51
Aristot. Pol. 1286a25-30; cf. 1287b30-7.
52
Aristot. Pol. 1268b5-10.
53
Aristot. Pol. 1298a1-35; cf. 1318b22-30.
54
Pace Bickford 1996; Mulgan 1999; Waldron 1999, pp. 92-123; Yack 1996.
55
Aristot. Pol. 1281a-b, 1286a25-32. In addition to the works just cited, see Goodin 2008, p.
94; Gutmann and Thompson 1996, p. 131, 2004, p.8; Landemore 2012, pp. 59-64; Ober 2008, pp.
110-14.
56
Cammack 2013a. Contra: Ober 2013, pp. 109, 111.
57
Aristot. Pol. 1281b30-35, 1282a614, 1282a2535, 1286a2528. Cf. Lane 2013.
58
Hansen 1983, pp. 119-21; Pirart 1974; Rhodes 1981.
59
Aristot. Pol. 1281b2082a16.
60
Aristot. Pol. 1253a117, 1332b5. Cf. Hist. Anim. 536b.
61
Cammack 2013a.
62
Aristot. Pol. 1298b20. Cf. Thuc. 1.87.4-5.

8
to balance those in the political class; and eliminating by lot any excess of the
former over the latter.63 The theme is balancing the numbers of notables and
members of the dmos who would take part in deliberation, and this suggests that
Aristotle was thinking principally of the outcome of the vote. What will apparent-
ly be better when more elite citizens take part is the final result, not (at least on
this evidence) the quality of any prior discussion. The same idea seems to have
underlain the custom of not allowing those with land near the frontier to take part
in deciding whether to go to war against a neighbouring polis, on the ground that
private interest would prevent them from being able to decide well (boule-
usasthai kals).64 Here, the use of the aorist confirms that Aristotle was thinking
specifically of the final decision.
According to Aristotle, then, bouleuomai meant coming to a decision about a
course of action within ones power, in political contexts perhaps especially
through voting, and the dominant paradigm was internal. Groups did deliberate,
but Aristotle did not specify exactly how. Evidently it did not strike him as suffi-
ciently importantwhich is in itself instructive.

Deliberation in early sources: the internal model confirmed

As noted above, Aristotles account of the purpose of deliberation is nowhere
contradicted. Where variations appear, they concern only the number of delibera-
tors and whether their activity was internal, dialogical, or a combination. None-
theless, the internal model came first, not only in Aristotles philosophical presen-
tation but also historically. It has been said that internal deliberation is modelled
upon our interpersonal experiences of discussion and debate, but the ancient
Greek evidence suggests otherwise.65 Every early use of bouleuomai, from Homer
to Sophocles, represents internal activity.
In its first appearance, in the Iliad, Zeus is said to have gone back on an earlier
promise and determined upon (bouleusato) cruel deceit.66 Theognis advises his
reader to think (bouleuou) twice and thricefor the headstrong man comes to
grief, while Semonides vicious woman all day longponders (bouleuetai) only
this, how to do the greatest harm she may.67 Later examples include Electras
Hear what I have determined (bebouleumai) to accomplish!; Oedipuss O Zeus,
what have you decreed (bebouleusai) for me?; and Phaedras statement that she
goes to die, but how shall be my own devising (bouleusomai).68
Such uses are just what Aristotle leads us to expect. But something else
emerges from our earliest texts that we might not have anticipated. Although
bouleuomai was always associated with internal decision-making, that activity
could also be represented by bouleu, plan. In Hesiod, Phocylides and Pindar,
only bouleu appears in this context, while in Homer and Aeschylus bouleuomai
appears only once, bouleu many times. In Sophocles and Euripides, however,

63
Aristot. Pol. 1298b10-28.
64
Aristot. Pol. 1330a22-6.
65
Goodin 2003, p. 169. Cf. Mulgan 1999, p. 195.
66
Hom. Il. 2.111-14, rpt. 9.18-21.
67
Theog. 633-4; Sem. Censure 81-2.
68
Soph. El. 947, Oed. Tyr. 738, trans. Lloyd-Jones; Eur. Hipp. 723, trans. Kovacs.

9
bouleuomai and bouleu are used about equally, while in the fourth century
bouleuomai was far more common. What accounts for these usage patterns?
The answer both confirms the association of bouleuomai with internal deci-
sion-making and, in introducing us to the wider family of terms related to plan-
ning and advising in the classical era, lays the groundwork for an improved un-
derstanding of the division of labour within and among Athenian political bodies.
We must here tackle another grammatical point. Technically, bouleu and
bouleuomai are not distinct verbs, but two voices of the same verb, bouleu the
active and bouleuomai the middle. The textbook distinction between these voices
is that the active represents the subject performing the action of the verb, while
the middle shows that the action is performed with special reference to the sub-
ject.69 Specifically, Smyth writes, as contrasted with the active, the middle lays
stress on the conscious activity, bodily or mental participation, of the agent.70
Heading Smyths examples are bouleu, plan, and bouleuomai, deliberate.
Drawing on their earliest uses, we may elaborate on Smyths distinction.
Though bouleu could show the subject planningthat is, designing and/or de-
ciding71his, her or their own action, it could also indicate the designing of a
plan by someone other than the decider of the action, i.e. the preparation or provi-
sion of advice. Bouleuomai, by contrast, emphasized that the subject of the verb
was himself or herself the decider. It indicated that the subject was engaged in
forming his or her own will (boul).
The flexibility of bouleu is clearly apparent in Homer. The subject is most
often a single person planning (or more negatively plotting) his or her own ac-
tion, as in Phoenixs report, Then I hatched a plan (bouleusa) to slay him with
the sharp sword.72 But plural subjects were also common. The Trojans wish to
know whether our foes are planning (bouleuousi) flight, so Dolon goes to Aga-
memnons ship, where the chief men will be holding council (boulas bouleuein,
literally planning plans).73 Here communication is certain, as in Achilles la-
ment, Never more in life will we sit apart from our friends and make plans to-
gether (boulasbouleusomen).74 Additionally, Nestors request that Agamemnon
follow whoever devises the wisest counsel (bouln bouleuese) and Odysseus
reference to a Phoenician who had given him lying counsel (bouleusas) suggest
advise, producing a plan for another rather than for oneself.75
In other archaic texts, we find not only plan, plot, and advise, but also
conspire, deliberate and consult.76 Several of these reappeared in Aeschylus.
Bouleu suggests plan or decide in the frantic exclamation of an Argive elder,
One who wants to act must first plan (bouleusai) what action to take; advise in

69
Smyth 1956, 1703, 1713.
70
Smyth 1956, 1728. Ancient Greek verbs also had a passive voice, which does not need to
be discussed here.
71
OED, s.v. plan.
72
Hom. Il. 9.458. Cf. 2.204-6, 9.96, 10.343-4; Od. 1.1443-4, 5.22, 9.299, 420, 11.229, 12.55,
14.490, 24.478. Plot: Od. 5.178, 187, 10.300, 344.
73
Hom. Il. 10.310-11, 325, tr. Murray, rev. Wyatt. Cf. 10.398, 14.464; Od. 23.217.
74
Hom. Il. 23.77-8. Cf. 1.531, 2.344-6, 377, 10.415; Od. 6.60, 13.439, 16.233-4.
75
Hom. Il. 9.74-5, Od. 14.295-7, trans. Murray, rev. Wyatt. Cf. Il. 10.146, 24.650.
76
Tyrt. Rhetra; Theog. 69-72, 1050-1, 1088; Pind. N. 9.37; Phoc. in Orion Anth. i.22.

10
Prometheus sad plaint, I, though advising (bouleuein) for the best, could not
persuade the Titans; and decide or decree in Eteocles declaration, If anyone
fails to obey my authority, a sentence of death will be decreed.77
Bouleuomai, by contrast, appeared only when the subject was himself or her-
self deciding, especially when this involved opposition to the will of another. In
the line from the Iliad quoted above, what will transpire is what Zeus has private-
ly determined upon, as opposed to what he had previously promised. Likewise,
Theognis Think twice and thrice and Semonides All day long she ponders
suggest private, even secretive, decision-making. Aeschylus sole use of the mid-
dle bouleuomai is especially revealing.78 When the Theban maidens defy Eteo-
cles, he tells them, in Sommersteins translation, not to behave imprudently
(bouleuou kaks).79 Behave is hardly a literal rendering of bouleou, but it aptly
conveys the connection between internal thought and the subjects own action.
The association between bouleuomai and deciding ones own action was then
reconfirmed in the striking reduction in the relative frequency of bouleu during
the classical period. While in Homer the proportion of active to total uses was
32/34, in the other archaic poets 8/10 and in Aeschylus 16/17, in Sophocles it was
13/22; Euripides, 18/32; Pseudo-Xenophon, 3/9; Aristophanes, 6/21; Herodotus,
23/133; Antiphon, 5/14; Andocides, 6/19; and Thucydides, 23/110. In Lysias, in
the fourth century, the figure was 23/42; Isocrates, 2/114; Xenophon, 9/114; Pla-
to, 5/103; Isaeus, 0/11; Aeschines, 6/46; Demosthenes, 14/168; Aristotle, 14/145;
Lycurgus, 0/10; Hypereides, 0/4; Dinarchus, 1/8; and Theophrastus, 0/3.
What accounts for the change is specialization. The early prominence of
bouleu gave way to less frequent usage as several of its functions devolved to
other terms. Most importantly, deliberate, i.e. come to a decision about a course
of action within ones own power became the near-exclusive province of
bouleuomai.80 What had been a way of emphasizing the subjects authority over
an action gradually became standard usage whenever the proposed action was de-
cided by the subject. Bouleu still performed this function in the fifth century81

77
Aesch. Ag. 1346, PB 169, trans. Sommerstein (modified), Seven 198, trans. Smyth. Cf. Ag.
1223, 1614, 1627, 1634; Eum. 696; Pers. 758; Seven 200, 248; PB 206, 1031.
78
Aesch. Lib. 716-18 and Ag. 844-5 are future middles used passively: see Smyth 1956,
807ff.
79
Aesch. Seven 223, tr. Sommerstein (modified). Cf. Aristoph. Eccl. 769-70.
80
Pace Ober 2013, p. 116, where bouleuein (the present infinitive active) is defined as to de-
liberate. See e.g. Soph. El. 1046, Oed. Tyr. 537, 1367, Ant. 772, Trach. 589, Phil. 1228; Eur.
Supp. 248, IA 1102, Ba. 842-3, Phoen. 735, Andr. 63, 1280, El. 269, Hipp. 900, Med. 567, 893,
Orest. 637-8, 1131; Aristoph. Frogs 865, Kn. 88, Lys. 951, Peace 58, 106, 230; Hdt. 1.20, 79, 91,
116, 3.17, 119, 134, 153, 154, 5.35, 111, 6.3, 86; 7.10, 12, 13, 49; 8.100, 101; 9.12, 13, 14, 16;
Thuc. 1.36, 2.64, 3.48, 5.8, 71, 6.12, 8.58; Antiph. 1.17; Andoc. 1.42, 145; Lys. 14.45; Xen. Hell.
1.30-1; Isaeus 1.11, 20, 41, 43, 50, 2.1,15, 3.64, 7.33, 8.36, 10.16; Isoc. 1.34, 35, 2.2, 47, 51, 3.51,
9.41; Plat. Rep. 390b, L1 309b, L7 324a, 346d-47c; Aeschin. 1.64, 2.114, 3.81, 209; Dem. 21.74;
Din. 2.24; Hyp. 6.15.
81
Soph. Oed. Tyr. 618, 1417, Ant. 1179; Eur. Ion 984, El. 618, Med. 874, Or. 773; Aristoph.
Birds 637, Clouds 419, Ass. 505, poss. Peace 690; Hdt. 1.73, 117, 120, 3.84, 6.52, 130, 8.97, 100,
9.106; Thuc. 1.85, 97, 132, 2.6, 3.28, 4.15, 37, 41, 51, 74, 5.63, 87, 111, 116, 6.18, 91, 8.50; Lys.
3.42; Dem. 19.21; Aristot. Met. 1013a.

11
and when the meaning was closer to plot than to plan.82 But eventually plot,
too, was largely taken over by another verb: epibouleu, literally against-plan,
first seen in Aeschylus.83 Similarly, and as we shall soon see significantly, ad-
vise was from the late fifth century nearly always expressed not by bouleu but
by symbouleu, literally with-plan, first seen in Sophocles.84
In place of these uses, bouleu became primarily associated with the activity
of formally constituted councils and councillors. It had, as we have seen, appeared
in this context in Homer; likely later examples appear in Herodotus, Pseudo-
Xenophon and Aristophanes, and definite examples in Antiphon and Andocides.85
Lysiass speeches, which date mainly from the early fourth century, contain only
one use of bouleu that does not indicate council activity.86 Those of Demosthe-
nes, by far the largest corpus among the extant orators, also contain only one.87
Conversely, in Demosthenes, come to a decision about a course of action within
ones power, without negative intent, was always represented by bouleuomai
just as in Aristotle.88
Demosthenes and Aristotle were exact contemporaries (384-22), so it may not
be surprising to find that they used bouleuomai in the same way. But we should
note how fully Aristotles analysis has been confirmed. Our earliest examples of
bouleuomai suggest not only that it originally indicated internal deliberation, but
that it actually emphasized the internality of the deliberation taking place. Later,
just as Aristotle attests, bouleuomai gradually became the only way to indicate
coming to a decision about the subjects own action. During the same period,
however, we also find our first examples of bouleuomai used with a plural sub-
ject; and many of these, unlike in Aristotle, featured group discussion.

Dialogical deliberation: decision-making through discussion

The earliest known cases of bouleuomai suggesting group discussion are mid-fifth
century: Have they in fact decided (bebouleuntai) to do this to me?, spoken by
Sophocles Electra with reference to her mother Clytemnestra and stepfather Ae-
gisthus, and Aeneas comment to Hector in Euripides Rhesus, if this signalling

82
Soph. Aj. 1055, El. 1001, 649, Oed. Tyr. 606, Ant.. 490, Trach. 807; Eur. Andr. 804, El. 27,
1012, Hec. 855, 870, Med. 37, 316, 401, Orest. 1089, Rh. 862, 950; Hdt. 1.11, 210, 5.106, 6.61,
7.197, 9.110; Antiph. 1.26, 3.7, 6.16; Andoc. 1.95, 2.20; Aeschin. 2.115, 117; Dem. 19.21; Din.
1.30; Aristot. Pol. 1310a10.
83
Aesch. Seven 29; Soph. Oed. Tyr. 618; Eur. Or. 1237; Hdt 1.24, 99, 183, 209, 210, 2.121,
3.119, 122, 134, 6.65, 137, 8.132; Aristoph. Kn. 894, Peace 404, 407, Thes. 83, 335, Plat. 570,
1111; frequently thereafter.
84
Bouleu suggests advise at Soph. Antiph. 267, Phi. 423, Eur. Hipp. 89, Rh. 108, Thuc.
3.42, 4.68, 6.39, 8.76, Lys. 31.31, and seldom thereafter. Symbouleu appears in Soph. Oed. Tyr.
1370; Hdt., 57 uses; Thuc. 1.65, 8.68; Aristoph. Cl. 475, 793-4, and frequently thereafter.
85
Hdt. 6.37; Ps. Xen. 1.9; Aristoph. Eccl. 444, Kn. 774, Peace 690; Antiph. 6.45; Andoc.
1.75, 90.
86
Lys. 3.42. Cf. 13.19, 20, 74; 16.8; 25.14; 26.10-11; 30.10, 22, 23; 31.1, 2, 5, 14, 24, 26, 32,
31, 34; Isoc. 16.43; Aeschin. 1.80, 110; 3.72.
87
Dem. 19.21 (suggesting plot). Cf. Aeschin. 2.115, 117. Contrast Dem. 18.25; 19.154, 286;
21.111; 22.5, 9, 12, 16, 36, 40; 47.44; 57.8; 59.3, 4.
88
Dem. 4.33, 8.67, 18.235, 272, 19.226, 21.41, 74, 23.12, 27.4, 36.31, 50, 37.13, 47.71,
52.30, 61.34, 41, 56, Ex. 19.

12
is a trap to catch us, we shalltake counsel (bouleusomestha).89 In neither case
is speech explicitly attested, though it may be assumed. Elsewhere it is certain.
Herodotus, for instance, used bouleuomai to refer to the first meeting of the
seven Persians who hoped to kill the Magian pretender to the Persian throne, a
meeting at which they explicitly exchanged speeches (edidsan sphi-
silogous).90 It reappeared in relation to the so-called constitutional debate said
to have taken place a few days later, when words were uttered which to some
Greeks seem incredible.91 Similar cases appear in Thucydides, such as the allies
deliberated (ebouleuonto) which of the remaining places they should go against
nextthe Eleans urged Lepreon, the Mantineans Tegea, and the Argives and
Athenians sided with the Mantineans;92 in Plato, as in Socrates comment to Cal-
licles, I once overheard you (pl.) debating (ebouleuonto) how far the cultivation
of wisdom should be carried, and the line in the Critias that ten kings took coun-
sel (ebouleuonto) about common affairs, agreeing that if anyone should attempt
to overthrow any citythey should all lend aid, taking counsel in common
(koinbouleuomenoi); 93 and in Demosthenes, the envoys met and discussed
(ebouleuonth) which of them should be left behind.94 What can we say about
these cases and others like them?
To begin with, each involves small numbers. Clytemnestra and Aegisthus are
two, the senior Trojans in Rhesus not many more, the Persian noblemen seven,
Callicles and friends four, the kings in the Critias ten, and the envoys in Demos-
thenes nine. The group in Thucydides may be larger, but not by much. Such fig-
ures are comparable to the eight to ten participants that make up focus groups to-
day or the fifteen to eighteen canvassed in deliberative polling.95 Then as now, it
would seem, group discussion coincided with limited participation.
Next, though bouleuomai in these examples certainly indicated discussion, so
could other verbs. One, already encountered in Aristotles report of the ban on
discussion among judges, was koinologeomai, speak together.96 Others included
didmi autous logous, give each other speeches, just seen in Herodotus; 97
anakoino, communicate, found in Lysistratas demand, get your allies heads
together (anakoinsate) and come to some decision (bouleusasthe);98 and di-
aleg, converse, source of the English dialogue. Dialeg often appeared in
philosophical contexts, and the speech it named, dialectic, was conventionally
contrasted with political speech, as in Platos Gorgias or the opening of Aristo-

89
Soph. El. 385, trans. Lloyd-Jones; Eur. Rh. 129-30, trans. Murray.
90
Hdt. 3.71. Cf. 3.74, 76, 84.
91
Hdt. 3.80. Cf. 1.112, 133, 143, 2.160, 4.102, 119, 137-8, 178, 5.92, 119-20, 6.7, 22, 138,
7.144-5, 173-5, 207-8, 219, 8.4, 7, 18, 40, 49, 57, 71, 75, 108, 130, 9.23, 41, 51-2, 86, 96-7, 106.
92
Thuc. 5.62. Cf. 6.1, 25, 46, 93, 7.1, 47, 50, 8.8, 54; Xen. Hell. 2.1.6, 31, 6.4.15.
93
Plat. Gorg. 487c, Criti. 119d, 120c-d, trans. Lamb. Cf. Meno 90e-91a, Charm. 176c, Laws
784b.
94
Dem. 19.122, tr. Vince. Cf. Lys. 13.24.
95
Fishkin 2009, p. 38.
96
Aristot. Pol. 1268b5-10. Cf. Thuc. 8.63, 98.
97
Hdt. 3.71, 76, 84, 6.138, 8.9. Cf. Eur. Orest. 774.
98
Aristoph. Lys. 1176, trans. Lindsay. Cf. Xen. Hell. 6.3.8; Isoc. 1.34, 5.19, 235, 12.235; Plat.
Prot. 314b, 349a; Aeschin. 2.64, 68.

13
tles Rhetoric.99 A more political example appears in a decree quoted in Demos-
thenes speech On the Crown, which directed envoys to visit Philip of Macedon
and confer (dialexontai) with him.100
Another option was koino, make common (middle koinoomai, make
common to each other). An important use appears in Aeschylus Agamemnon.101
Hearing Agamemnons groans, the chorus of elders suspects he has been killed
and, in Sommersteins translation, declares Let us deliberate (koinsmeth) and
see if there might be any safe plan to follow. What happens next looks like ideal
deliberative practice on the dialogical conception. A small number of speakers
(between six and twelve) discuss what the group should do, responding to each
other and giving reasons for their positions. Three proposals are made: to call for
help from the citizenry, to apprehend the murderers on the spot, and to ascertain
the facts before proceeding. The last secures general assent and they act accord-
ingly. This is probably the best ancient Greek example of dialogical deliberation,
but it is represented not by bouleuomai but by koinoomai, a verb never seen in a
mass political setting.102
What distinguished cases where bouleuomai was used from those where it
was not? The principal difference was exactly what Aristotle leads us to expect.
Bouleuomai appeared only when what was reported was not merely discussion,
but specifically coming to a decision about a course of action within the groups
powerand sometimes not even then, as the Agamemnon example shows. Hero-
dotus constitutional debate is known for its theoretical content, but the reported
context was practical. The future government of the Persians lay in these mens
hands, and after three speeches, they voted, with four out of the seven favouring
monarchy.103 Similarly, though a conversation on the cultivation of wisdom may
sound fully philosophical, the exhortations of Callicles and his friends to each
other to beware of making yourselves overwise imply a practical interest: they
are deciding what style of life to pursue.104 The allies next steps, the ten kings
policymaking and the envoys plan also fall into this category. Conversely, when
only discussion, not decision-making, is intendedas in the decree ordering the
conference with Philip, after which the envoys were to report back to the assem-
blya different verb is used.
Most importantly, when bouleuomai indicated dialogical deliberation, every
member of the group was involved in making the decision. This is clearest in the
Persian example, since the debate culminates in a formal vote. It is also evident in
Thucydides: the Mantineans plan is adopted because it is supported by the Ar-
gives and Athenians, leaving the Eleans in a minority. How the council of kings in
the Critias reaches its decisions is not specified, but we can assume that no one
attended simply to air his opinion. Each king, as sovereign in his own community,

99
See especially Plat. Gorg. 448e-49b, 500c; Aristot. Rh. 1354a1. Cf. Isoc. 5.18, 234, 9.34;
Plat. Prot. 314c-d, 335d-36a. For discussion, see Chambers 2009; Garsten 2006.
100
Dem. 18.164, cf. 28, 73; Isoc. 2.46, 15.256; Aeschin. 2.18, 103.
101
Aesch. Ag. 1347-71, cf. Lib. 673, 716-18; Aristoph. Cl. 197; Thuc. 4.4.
102
See also the soldiers assembly at Thuc. 8.63, 76-7; again, bouleuomai does not appear.
103
Hdt. 3.83.
104
Plat. Gorg. 487d, trans. Lamb.

14
would be expected to have a vote, as well.105 The same seems true of Callicles
and friends and the envoys in Demosthenes. The decisions they are making
respectively, how far they will pursue wisdom and who among them should be
left behindare ones in which all can be expected to participate, not merely
through their voices but also through their votes.
In Aristotelian terms, bouleuomai was used in discursive contexts when each
member of the group took part not only in considering, but also in deciding what
to do. As we have seen, the decision was often taken by a vote; establishing a
consensus verbally would also have been possible. Either way, the deliberative
process culminated in the choice of an action for which the deliberating agent,
here the group as a whole, was responsiblejust as in previous examples.

Guided deliberation: internal decision-making with advice

The third type of deliberation found in our ancient Greek sources is a partial com-
bination of the first two. Like dialogical deliberation, it involved communication,
yet the deliberation that took place was internal. Several people might offer opin-
ions, but only one agentoften a single personmade the decision, and to that
agent only was bouleuomai applied.
We may call this model guided deliberation, since those who deliberated were
guided by the advice of others. Typical cases involved a ruler and his counsellors,
such as the Milesian tyrant Aristagoras and his followers prior to the Ionian re-
volt, Xerxes and his commanders before his return to Persia, or Philip of Macedon
and his would-be advisor Isocrates.106 Other examples involve a father taking ad-
vice from his sons and a husband from his wife.107 In each case the decision-
maker received input from others, often serially rather than through group discus-
sion, but the final decision was his alone.
Two consecutive speeches in the Demosthenic corpus illustrate the difference
between this and dialogical deliberation. In Against Olympiodorus, two men
swear to proceed by mutual agreement and take counsel together several times
before relations go sour. Here bouleuomai appears in the plural and koin, in
common, is added for clarity. Mutual agreement is koin bouleuomenoi (a pre-
sent participle) and the two mens discussions are described by ebouleuometha
koin, we came to a decision together.108 This is a clear example of dialogical
deliberation: joint decision-making through discussion. In Against Evergus,
however, though discussion takes place, the decisionand thus the attribution of
bouleuomai, deliberatefalls to one man only. The prosecutor describes delib-
erating with his friends, but, crucially, bouleuomai appears in the singular:
ebouleuomn meta tn philn, I deliberated with my friends.109 Though others
spoke, only one deliberated, because only one decideda clear example of guid-
ed deliberation.

105
Schwartzberg 2010.
106
Hdt. 5.36, 8.101, Isoc. 5.18 (quoted below).
107
Hdt. 1.61, 2.107.
108
Dem. 48.9, 10, 22, 28. Cf. Lys. 13.24; Isoc. 17.7; Hyp. 3.12.
109
Dem. 47.71. Cf. Thuc. 1.128; Isoc. 5.69, 12.233; Plat. Rep. 400b.

15
 What verb, if not bouleuomai, represented the activity of those who spoke but
did not decide? The answer is symbouleu, advise or counsel, as encountered
above, or sometimes a synonym.110 While the speaker in Against Evergus delib-
erated, ebouleuomn, his friends advised, symbouleuontn.111 Similarly, Xerxes
asked Artemisia to advise (symbouleuson) me as to which of these things I shall
best decide (bouleusamenos) to do,112 and Isocrates friends cautioned him, You
are about to send something offering advice (symbouleusonta) to Philip, a man
whocan hardly fail to believe that he more than anyone is able to deliberate
(bouleuesthai) by himself!113
That bouleuomai and symbouleu often appeared together should be no sur-
prise. As discussed above, both denoted meanings at one time expressed by
bouleu: they represented two strands of a single advising/deciding dyad. In
Greek, this impression is reinforced by the fact that the middle and passive forms
of the Greek verb look the same. Bouleuomai could mean either come to a deci-
sion or be advised. The English equivalent is take counsel: one may take
counsel either alone or more literally from others. Bouleuomai appears in the lat-
ter sense in Aeschylus Libation-bearers: Since we are not short of friends, Cly-
temnestra announces, we will take counsel (bouleusomestha).114 The mention of
friends makes it clear that advice is expected. This situation could also have been
expressed by the middle voice of symbouleu, i.e. symbouleuomai, consult, as
seen several times in Herodotus. When Sesostris house is set on fire, he at once
consulted his wife (symbouleuesthai t gynaik), that is, took advice from his
wife. 115 Similarly, Masistes, in a moment of uncertainty, symbouleusamenos
(sing.) toisi paisi, consulted his children.116 These cases are equivalent to that
found in Against Evergus. Though discussion took place, the use of the singular
shows that Sesostris and Masistes deliberated, i.e. decided, alone. They were the
rulers in these contexts, their wives or children their counsellors.

II. Deliberation in Classical Athenian Political Settings

Three models of deliberation thus appear in our ancient Greek sources: one inter-
nal, one dialogical, and one guided. In each case, the deliberator or deliberators
came to a decision about a course of action within his, her, or their own power.
How did the activities of the Athenian courts, council, and assembly compare?

The courts: the internal model aggregated

110
E.g. peithomai, persuade, Hdt. 1.124; paraine, exhort, recommend, advise, Hdt. 9.79,
Isoc. 2.46, L3.3.
111
Dem. 47.71.
112
Hdt. 8.101.
113
Isoc. 5.18. Cf. Hdt. 3.156, 5.124, 7.10; Xen. Hell. 2.14-15; Isoc. 1.5, 34-5, 44, 2.6, 49-53,
3.8, 12, 17, 5.88, 6.4-5, 8.13, 54-5, 12.170-1, 17.7-8. Cf. Aristot. Nic. Eth. 1112b8-12, Lys. 22.8-9.
114
Aesch. Lib. 718 (a future middle used passively). Cf. Ag. 844.
115
Hdt. 2.107.
116
Hdt. 9.113. Cf. Aristoph. Cl. 457; Thuc. 8.68; Xen. Cyrop. 2.1.7; Plat. Theag. 122a; Isoc.
9.44.

16
In Athens courts, charges were brought by volunteer prosecutors who were given
a fixed amount of time to make their case. The defendant had the same amount of
time to respond, and a panel of citizen-judges, selected daily by lot from among
volunteers (two hundred or more for dikai or private charges, five hundred or
more for graphai or public ones) voted their decision.117
As noted above, these judges certainly deliberated. Do not discover too late
that you have put to death an innocent man, pleaded a client of Antiphon. Rather
decide (bouleusasthe) well while there is still time.118 Lycurgus, opening a pros-
ecution for treason, urged: You who in your deliberations (bouleuomenous) are
now defending your fathers, wives, and childrenbe inexorable judges.119 Ae-
schines, prosecuting Demosthenes in On the Crown, asked his audience to de-
liberate (bouleuesthe)not as for some foreign state, but for your owndecide
(bouleusasthe) not with the help of your ears alone, but with your eyes, looking
sharply among yourselves to see whoproposes to aid Demosthenes.120 Dinar-
chus, also prosecuting Demosthenes a few years later, told his listeners to re-
member these things and decide (bouleuesthe) wisely, and asked, will you
preserve him? Not if you deliberate (bouleuesthe) well.121
Some of these exhortations are present tense, suggesting the process of con-
sideration; others are aorist, specifying the moment of decision. All are second
person plural imperatives, paralleled by other requests to think the matter through
carefully.122 And the form of deliberation they imply is internal. Modern judicial
deliberation, in which groups of a dozen or fewer judges or jurors discuss cases
and decide them either by consensus or by majority vote, is certainly dialogical.123
But as noted above, Greek judges, including Athenian ones, did not normally dis-
cuss cases. Although shouting and brief conversations were possible, judges were
not expected to communicate with one another and each mans vote was secret.
Indeed, judicial deliberation in Athens was not even precisely collective.
Votes were cast serially, and the collective quality of the decision came solely
from their final aggregation, which took place after both the consideration and
decision-making stages were complete. Perhaps because of this, judicial panels
were never addressed by the collective singular dikastrion, court, but always in
the second person plural, often as andres dikastai, men of the jury (literally
judge-men.124 Altogether, deliberation in Athenian courts resembled an aggre-
gated version of the internal model: internal decision-making by a large number
of single agents.

117
Boegehold 1995, pp. 21-42.
118
Antiph. 5.71, trans. Maidment; cf. 5.73, 90-1, 94.
119
Lyc. 1.2, trans. Burrt; cf. 1.11, 14-15, 83.
120
Aeschin. 3.255, trans. Maidment (modified).
121
Din. 1.26, 98, trans. Burrt. Cf. Andoc. 4.7; Lys. 6.8, 9.15, 19, 21.13, 25.21, 23, 28.16; Isoc.
5.170, 178, 248; Dem. 20.15, 35, 25.14, 48.52, 53.29; Aristot. Pol. 1286a25-30, 1287b30-5; Hyp.
5, fr. 3.
122
E.g. skepsasthe, consider, Dem. 21.73, 25.73, Lyc. 1.52; skopeite, examine, Dem.
20.36, 19.105, 24.117, 25.38, 35.54, Din. 1.111; logizesthe, reflect on, Dem. 22.22; enthymeis-
the, think, Dem. 19.310, 21.197; thersate, contemplate, Lyc. 1.75, 111, Din. 1.26, 75.
123
Abramson 1993; Allen 2000, pp. 3-11.
124
E.g. Lys. 13.92; Aeschin. 1.78; Lyc. 1.80.

17
The council: not normally a deliberating body
The Athenian council (boul) was a body of five hundred citizens responsible for
the daily administration of the city, including overseeing offices, welcoming offi-
cial visitors and, most important, preparing material for assembly meetings. This
often involved approving draft decrees, each one attributed to a named council-
lor.125 Like a judicial panel, the council was chosen by lot from volunteers; unlike
a judicial panel, it was constituted not for a single day, but for a year, meeting
most days.126 Councillors worked especially closely with other members of their
tribe (one of ten), even living and dining together when that tribe held the presi-
dency.127 These conditions would seem to favour deliberation on the dialogical
model, as several scholars have suggested.128 What can we tell from the evidence
available?129
In our sources, the Athenian council is the subject of bouleuomai eight times.
In three cases the council was acting as a court, as it occasionally did; as such, we
must assume that normal judicial procedure was followed, i.e. no discussion
among the judges (we know for sure that a secret ballot was used).130 Two cases
concern the oligarchical council of the future designed by conspirators in 411
but never realized: we cannot treat these as evidence of the normal activity of the
democratic council.131 Two more are from Lycurgus, in a speech referring to the
time of Athens greatest crisis, its defeat by Philip of Macedon at Chaeronea in
338.132 Only Pseudo-Xenophons text, the council has to consider (bouleuesthai)
many issueswar, revenues, law-making, things do with the polis, allies, tribute,
care of dockyards and shrines, uses bouleuomai in connection with the normal
running of the democratic council, and he is a notoriously sloppy writer.133
However, we find scores of examples of bouleu, plan or, as noted above,
perform council activity,134 and probouleu, pre-plan, used in connection with
the council.135 And this is, in fact, just what we should expect. Bouleuomai indi-
cated that the subject was coming to a decision on the matter under consideration,
but the council almost never decided the issues that came before it.

125
Ps. Aristot. Ath. Pol. 43.2-4, cf. Aristot. Pol. 1299b30. Rhodes 1972, pp. 52-87.
126
Ps. Aristot. Ath. Pol. 43.3; Rhodes 1972, p. 30; Hansen 1999, pp. 250-1.
127
Rhodes 1972, p. 16.
128
Fishkin 2009, pp. 11, 82; Ober 1989, pp. 142-51; Ober 2013; Ruz 1997, pp. 437-543.
129
Illustrating the relative lack of evidence, the accounts in Ober 2008 and 2013 are hypothet-
ical.
130
Lys. 26.21, 24; Dem. 51.9. Rhodes 1972, pp. 38-9.
131
Ps. Aristot. Ath. Pol. 30.4.
132
Lyc. 1.37, 126.
133
Ps. Xen. Ath. Pol. 3.2; Marr and Rhodes 2008, p. 8.
134
Ps. Xen. 1.6, 9; Aristoph. Ass. 444, Kn. 774; Antiph. 6.45; Andoc. 1.75, 90, 91, 95; Lys.
13.19, 20, 74, 16.8, 25.14, 26.10-11, 30.10, 22, 23, 31.2, 5, 14, 24, 26, 31, 32, 34; Plat. Ap. 32b,
Gorg. 473e; Aeschin. 1.80, 110, 3.76; Dem. 18.25, 28, 19.154, 286, 21.111, 22.5, 9, 12, 16, 36, 40,
47.44, 57.8, 59.3, 4; Aristot. Pol. 1282a32, 1306b6; Ps. Aristot. Ath. Pol. 4.3, 30.2-3, 30.6, 31.1, 3,
45.3, 62.3. Cf. Xen. Hell. 1.31.
135
Thuc 8.1.2; Dem. 18.53, 169, 19.34, 185, 20.4, 21.162; Aeschin. 2.58; Ps. Aristot. Ath.
Pol. 45.4. Cf. Hdt. 1.133; Aristot. Pol. 1298b30, 1299b30, 1322b16.

18
 When acting as a court, it did: thus Lysias and Demosthenes uses of
bouleuomai with respect to councillors acting as judges was appropriate. Equally,
oligarchical councils made decisions, as did democratic councils in emergencies,
such as when (as in Lycurguss speech) the citizenry was under arms.136 But un-
der normal circumstances, only the Athenian assembly and courts made decisions
on behalf of the polis. Even when a decree was originally prepared by a council-
lor, the assembly could change or reject anything it disliked, and this authority
was reflected in Athenian political terminology. As early as Homer, as we saw
above, bouleu had been used to describe the production of plans by agents who
were not themselves decision-makers (leading to the usage advise).137 This was
exactly the role of the council in democratic Athens. Indeed, it was the role of
both the council and the assembly in 411, when the polis was temporarily gov-
erned by a gang of oligarchs, and Thucydides, accordingly, used bouleu to refer
to the activity of both those bodies at that time.138 On the ancient Greek concep-
tion, however, this activity was not deliberation but planning or preparation.
Was such planning nonetheless dialogical? Even that much is not clear. The
acoustics of the council-chamber were good, and a defence speech written by An-
tiphon for a former councillor suggests that voicingopinions was a regular
part of a councillors role.139 Andocides also claimed that councillors, unlike as-
semblygoers, had the opportunity to consider (skepsasthai) issues at leisure,
which may suggest prolonged discussion.140
On the other hand, five hundred participants seems too many for genuine
group discussion, and what we know of proceedings tends to support that view.
Meetings followed an agenda, set by the tribe in presidency, and speakers spoke
not conversationally, from their seats, but serially, from a central platform: both
features imply oratorical rather than dialogical behaviour.141 Written submissions
were also used. The Greek word propose, graph, literally meant write, and
the council retained a secretary to read proposals and other documents aloud and
to assist in drafting when necessary.142 There is no evidence that submitting a
proposal necessitated making a speech: indeed, the common description of politi-
cal activity as speaking and proposing (legein kai graphein) shows that the two
were regarded as distinct tasks.143 Most significantly, Demosthenes and Aeschines
both distinguished the small number of talkers (legontes) or orators (rhtores)
on the council from the rest; Demosthenes adds that the latter for the most part do
not even go to the council-chamber.144 Shouting or uproar (thorubos) seems to
have been a more normal way for the majority to make its voices heard.145

136
Lyc. 1.37.
137
Hom. Il. 9.74-5, 10.146, 24.650; Od. 14.295-7.
138
Thuc. 8.66.1, cf. 72.1.
139
Johnstone 1997, pp. 105-6; Antiph. 6.45.
140
Andoc. 2.19.
141
Rhodes 1972, pp. 20, 30, 37-8, 80; Johnstone 1997, p. 105.
142
Dem 18.25, 19.154; Aristoph. Thesm. 375, 431-2; Rhodes 1972, pp. 134-41.
143
Dem. 18.66, 86, 88, 173, 302, 307, 19.35, 22.16, 30, 32-3, 23.3, 146; 26.23; Din. 1.100,
2.12; Ps. Aristot. Ath. Pol. 39, 45.
144
Dem. 22.36-8. Cf. Plat. Laws 758b; Rhodes 1972, p. 39.
145
Aeschin. 3.9; Andoc. 1.43-5, 2.15.

19
 The regular activity of the council was thus not only not called deliberation,
represented by bouleuomai, it did not even necessarily resemble group discussion
of the kind found in Homeric councils, represented by bouleu. What about in-
formal political discussion, for example at dinners? Presumably there was a lot of
political conversation, but since it will have been even more divorced from deci-
sion-making than councillors speeches during meetings, we may assume that
bouleuomai was not used.

The assembly: mass guided deliberation

Since at least the era of John Stuart Mill, the Athenian assembly has often been
called the Ecclesia.146 But ekklsia meant meeting (literally call-out) and was
virtually never the subject of a verb. The assembly as an agent was represented by
dmos; hence dmos, implying the assembly, was often the subject of bouleuomai
in decrees, speeches and in the philosophers and historians.147 What then did ho
dmos ebouleuse, the assembly deliberated, mean?
Many translations suggest dialogical deliberation. Adams Aeschines cites a
motion restricting the dmoss discussion (bouleusasthai) of peace to particular
days, and another specifying that the dmos should discuss (bouleusasthai) an
alliance.148 Maidments Andocides recalls how after the rule of the Thirty Ty-
rants you discussed (ebouleusasthe) ways and means of reuniting the city and
called a meeting to discuss (ebouleusasthe) revising the laws.149 Vinces De-
mosthenes berates his audience for not discussing (bouleuesthai) any question at
your leisure, but waiting until youre already losing, and elsewhere argues that
while the sum of money youre discussing (bouleuesthe) is a trifle, the habit of
mind it fosters is serious.150 The translation debate has a similar effect.151 Like
deliberate it may suggest either internal activity or dialogue, but in this context
dialogue is likely to be assumed.
As noted above, however, with five to eight thousand attendees, later fourteen
or even twenty thousand, genuine group discussion was impossible. Not only was
the group far too large, meetings, at two or three hours long, were too short.152
The vast majority of assemblygoers at any given meeting will thus not have been
able to speak.153 And most, surely, will not have wanted to. The formal barrier to
speaking was certainly low: one only had to approach the stage at the heralds
call, who wishes to speak (tis agoreuein bouletai)?154 But not everyone relishes
the prospect of addressing an audience of thousands, especially one given to
146
E.g. Mill 1978 [1856], p. 326, following Grote 1846-56.
147
Hansen 2010; Aeschin. 2.60, 67-8, 110, 3.71; Dem. 18.73, 165, 20.4, L1.2; Din. 1.90; Ps.
Xen. Ath. Pol. 1.16, 2.17; Plat. Prot. 319a-d, Gorg. 500c, 515d; Aristot. Pol. 1282a25-35,
1298b20-35; Thphr. Char. 26.
148
Aeschin. 2.109-10. Cf. 2.134, 3.71.
149
Andoc. 1.73, 82. Cf. 2.19, 3.40.
150
Dem. 10.29, 13.2. Cf. 15.6, 18.24, 19.13, 234.
151
Dem. 10.10, 15.1, 31, 25.9, trans. Vince; Isoc. 13.2, trans. Norlin; Thphr. Char. 26.1.
152
Aristot. Rh. 1354b1-4; Cicero Pro Flacco 15ff. Hansen 1999, pp. 136-7.
153
See n. 17 above.
154
Aeschin. 1.23, 27, 3.2-4, 20; Dem. 1.1, 18.170; Plut. Cim. 8.1; Lewis 1971, pp. 136-7.
Bouletai, wishes or wants, is distinct from bouleuomai, deliberate: boulomai refers to the
formed will, bouleuomai to the development of the will.

20
shouting and abuse.155 The acoustics were also difficult. The Pnyx, where meet-
ings were usually held, was simply a hollow on an open hillside: reconstructions
suggest that one would had to have been a trained speaker in order to be heard,
and even then, Christopher Johnstone argues, during the fifth century, it is doubt-
ful whether even half of the 5000 present could regularly understand what speak-
ers were saying.156 Things improved after 403/2, when the 180 reorientation of
the space (requiring the construction of enormous new retaining walls) blocked
some of the wind.157 But Aeschines strong voice still provoked envy, if Demos-
theness jibes are any guide; Demosthenes himself trained for months in order to
be heard; and Isocrates never spoke in the assembly, blaming his weak voice.158
Perhaps most revealing, the position of secretary to the assembly, responsible for
reading motions and other documents aloud (a position once held by Aeschines),
was, unusually for Athens, elected.159 This cannot have been necessitated by une-
ven levels of literacy, since other positions requiring literacy were filled by lot.160
Rather, it must have been sufficiently audible speech that was in short supply.
Yet if genuine group discussion must be ruled out, what we may call vicarious
group discussion need not be. The entire gathering may have been conceived as
engaging in dialogical deliberation via the contributions of some few speakers.
Exactly how few is debatable: Hansen estimates that 700-1400 citizens (some 2-
5% of the citizen body) acted occasionally as rhtores, orators or politi-
cians,161 in the years 355-22, in addition to the ten or twenty at any given time
who spoke regularly.162 But Hansens calculations are based on the number of
proposers and amenders found in extant decrees, and it is not clear that all or even
most of these men actually spoke in the assembly.163 As noted above, many de-
crees were initially drafted by councillors, and an oral defence before the dmos
does not seem to have been required.164 Indeed, Hansen himself has argued con-
vincingly that many decrees were passed without debate, through procheirotonia,
literally early raising of hands.165 Moreover, even new proposals or amendments
made from the assembly floor could be submitted in writing, read aloud by the

155
Aristoph. Ass. 86-7; Dem. 2.29, 19.23; Hyp. 5, col. 12. Cf. Xen. Mem. 3.6 and Plat. Alc. I;
Thuc. 1.42, 4.27-8; Eur. Orest. 901-2; Aristoph. Ach. 37-9; Lys. 12.73; Aeschin. 1.80-5, 2.84;
Dem. 13.3, 19.122, Ex. 4, 5, 21, 26, 47, 56; Rhodes 2016, pp. 248-50; Roisman 2004;
Schwartzberg 2010; Tacon 2001; Villacque 2012; Wallace 2004.
156
Johnstone 1997; cf. Ober 1989, p. 138. Aristot. Rh. 1414a16-17. Note also the portrayal of
Demos as half-deaf at Aristoph. Kn. 42.
157
Forsn and Stanton eds. 1996.
158
Dem. 19.206, 216, 338-9; Plut. Dem. 6-11; Isoc. 12.9, L8.7.
159
Ps. Aristot. Ath. Pol. 54.5; Dem. 19.70.
160
Ps. Aristot. Ath. Pol. 54.3-4.
161
Hansen 1989, pp. 1-24, avoids politicians, but cf. Ober 1989, p. 106, with whom I agree.
See also Connor 1972, pp. 116-17.
162
Hansen 1989, pp. 93-127, 1999, p. 272. Others, such as Ober 1989, pp. 107-9, prefer a
higher estimate, but as Hansen 1989, p. 124, emphasized, that entails a significant decrease in the
number of occasional rhtores.
163
Hansen 1989, pp. 97-8, 124-5, with Aeschin. 3.125, 159; Dem. 59.43.
164
Ps. Aristot. Ath. Pol. 29.1, 45; Hansen 1989, p. 97, with notes; Rhodes 1972, pp. 52-87, cf.
2016, pp. 247-8.
165
Hansen 1989, pp. 123-30. Cf. Aeschin. 1.23; Ps. Arisot. Ath. Pol. 43.6; Dem. 24.11.

21
secretary, and immediately put to the vote.166 Some, no doubt, will have provoked
debate, but many need not have required it: for example, the request that a refer-
ence to Skiathos in a decree be corrected to Old Skiathos, that certain Samians
be given a dinner, or that a third brother be honoured alongside two others.167
The number of proposers and amenders of decrees is thus not a good guide to
the number of assembly speakers. Yet even if the overwhelming majority of at-
tendees only listened and voted, the entire group may still have been conceived as
engaged in discussion, especially since all had the right to speak. Something like
this is often implied by ancient historians as well as political theorists. The Athe-
nians are said to have had an ideal of mass discussion, even if it was never
achieved: the relevant distinction is said to have been between active partici-
pants, who spoke, and passive participants, who listened and voted.168 Was this
how the Athenians saw it?
One point suggests that it was: the use of bouleuomai in the first person plural
by speakers to refer to both speakers and listeners. Diodotus, for example, when
attacking Cleons proposal regarding the Mytilenaeans, argued: We are deliberat-
ing (hmasbouleuesthai) not about the present but about the futurewe are
considering (bouleuometha) a question of policy.169 Similarly, Aeschines argued
that Solon had laid down the proper manner of conducting our deliberations
(hmasbouleuesthai); and Isocrates and Demosthenes also used these forms.170
Such uses of the first person plural by speakers are expected in modern delib-
erative settings and have sometimes even been required.171 But in the classical
Athenian assembly they were in fact rare. They usually appeared when speakers
were emphasizing their identification with their audience, as when distinguishing
themselves from a rival speaker (Diodotus), showing respect to a founding hero
(Aeschines), or discussing an external threat to the polis (Demosthenes).172 At
other times, speakers commonly used the first person plural to refer to themselves
and other speakers, as in Demosthenes reference to all of us who address you,
or Hypereides use of it to refer to the dmos and us to those who spoke before
it.173 Most significantly, the vast majority of uses of bouleuomai by assembly
speakers were in the second person plural, referring to the activity of the audience
as opposed to that of the speaker.
This may have been noticed in the lines from Andocides and Demosthenes
quoted at the beginning of this section. But it extended far more widely. Cleon
told his audience, You (hymeis) are more like men at an exhibition of sophists
than deliberators (bouleuomenois), and other Thucydidean orators (in Athens and
elsewhere) spoke similarly.174 Dinarchus described the herald praying before he

166
Aeschin. 2.64, 68, 83-4.
167
IG i3 110, IG ii2 1, 212.
168
Hansen 1989, pp. 11-17, 1999, pp. 267-8; Manin 1997, p. 16; Ober 1989, pp. 324-5;
Rhodes 1995, p. 159, though cf. 2016, p. 264; Urbinati 2000, pp. 762-3.
169
Thuc. 3.44.
170
Aeschin. 1.22, 2.62, trans. Adams. Cf. 1.33, 3.68. Isoc. 8.15, 52. Dem. 23.211, 24.99.
171
Gutmann and Thompson 2004, p. 18 (quoting the Oregon Health Commission).
172
Thuc. 1.86, 3.44; Andoc. 24, 29, 34; Dem. 6.5, 8.3, 9.7, 17.17, 18.89, 196.
173
Dem. 14.2; Hyp. 5, cols. 28-9. Cf. Thuc. 3.37-8, 43; Andoc. 3.40; Isoc. 13.8; Dem. 8.22.
174
Thuc. 3.37-8. Cf. 1.72, 73, 78, 80, 4.87, 6.17, 6.36, 92, 7.14.

22
hands over to you the task of deliberation (hymin to bouleuesthai), while Aeschi-
nes reported Demosthenes saying to the assembly, You have your answer; it re-
mains for you to deliberate (hyminbouleusasthai).175 Other notable lines in-
clude these speakersmake the mistake of submitting to you the wrong subject
for deliberation (bouleuesthe); if in the past their advice had been sound, there
would have been no need for you to deliberate (hymasbouleuesthai) today; and
deliberation (to bouleuesthai) is naturally difficult, but youhave enhanced the
difficulties; others deliberate before the event, you (hymeis) after.176
Deliberation was thus typically represented as the task of the audience. But
then what did speakers do? The answer recalls the guided type of deliberation de-
scribed above. While the audience deliberated, represented by bouleuomai, speak-
ers advised, represented by symbouleu.177
When they have to decide (bouleusasthai) something to do withthe polis,
the man who rises to advise (symbouleuei) themmay equally well be a smith, a
shoemaker, a merchant, a naval captain, rich, poor, well-born or otherwise,
claims Socrates in the Protagoras, using a formulation found throughout our
texts.178 Demosthenes, according to Aeschines, declared that he was amazed at
both parties, as well the listeners as the ambassadors, for they were carelessly
wasting timethe listeners the time for taking counsel (bouleuesthai), the ambas-
sadors the time for giving it (symbouleuein).179 Demosthenes also said that it was
your duty when deliberating (bouleuomenous)to be willing to listen to all your
advisers (tn symbouleuontn); that when considering (bouleuomenous) such
important matters, you ought to give each of your advisors (tn symbouleuontn)
freedom of speech; and that while all previous advisers (tn symbebouleukotn)
were inadequate, he would explain what would profit you who deliberate (tois
bouleuomenois). Many further examples can be found.180
A line from Aristotles Rhetoric illustrates both the difference between
bouleuomai and symbouleu and the way that that difference may be effaced in
translation. One rendering refers to the most important subjects of deliberation,
and those most often discussed by deliberative speakers; another has about
which all men deliberate and deliberative orators harangue.181 A more accurate
rendering would be about which all men deliberate (bouleuontai) and those who
advise (hoi symbouleuontes) speak publicly (agoreusin). In the Greek, if not the
English, deliberating and advising were distinct.

175
Din. 1.14, trans. Burrt; Aeschin. 2.50, trans. Adams. Cf. Andoc. 1.73, 75, 2.19, 3.12, 33;
Aeschin. 2.60, 61, 70, 82, 134; 3.69, 120, 150-1, 251-2.
176
Dem. 3.1, 4.1, 5.2, trans. Vince. Cf. 4.33, 41, 5.3, 6.28, 8.1, 9.20, 10.1, 28-30, 13.2, 15.1-2,
6, 13, 21, 31, 18.24, 65, 86, 19.5, 13, 34, 96, 206, 23.109, 115, 24.32.
177
Kallet-Marx 1994; Ober 1989, pp. 317-23, 2008, p. 164.
178
Plat. Prot. 319b-d. Cf. Prot. 322d-24c; Gorg. 455b-56a; Alc. I 106c-7d; Xen. Mem. 3.6;
Hell. 1.7.16, 19, 2.2.15, 2.4.40; Andoc. 4.12; Lys. 25.27, 33.3; Isoc. 4.3, 19, 170-1, 5.143, 8.1, 27;
Aeschin. 1.26, 64, 1.110-11, 120, 180, 186, 2.29, 65, 157, 165, 3.158, 225-6; Dem. 8.73, Ex. 3, 33;
Din. 1.31, 35-6, 40, 72, 76-8, 81, 93, 2.14, 15; Hyp. 5 col. 28; Ps. Aristot. Ath. Pol. 23-4, 29.
179
Aeschin. 2.49, trans. Adams.
180
Dem. Ex. 26-7, 6. Cf. 4.1, 8.1, 10.17, 75, 14.8, 15.1, 18.86, Ex. 1, 11, 20, 27, 30, 33, 35-6,
56; Thuc. 3.43, 38; Lys. 6.54; Isoc. 3.19, 21, 5.88, 8.1-2, 8-13, 52-55, 7.77-8, 12.170-1, 15.256.
181
Aristot. Rh. 1359b20, trans. Lawson-Tancred, Freese.

23
 This distinction should not be overstated. There was an important connection
between deliberating and advising, as Aristotles claim that all men deliberate
implies. Yet this connection only underlines the difference between deliberators
and advisors at the moment of communication. Speakers did not deliberate along
with the audience not because they never deliberated, but because they had al-
ready done so. The audience was supposed to have an open mind during meet-
ings: as Demosthenes argued, sound judgment meant not having decided (be-
bouleuesthai) before you have heard that upon which you should base your deci-
sion (bouleusasthai).182 This entailed giving all speakers a hearing and listening
quietly.183 But the point of listening quietly was to get the benefit of each speak-
ers prior thought. The same arguments which we use in persuading others when
we speak in public, we employ also when we deliberate (bouleuomenoi), Isocra-
tes explained.184 Aeschines attributed both his speeches and his silences to hav-
ing deliberated (bouleusamenos),185 while Demosthenes reported that it is diffi-
cultnot only to say before you what must be done, but even to have found it out
by solitary reflection. Good policies were rare and hard to discover; still he
wished to tell you what I have convinced myself is expedient.186
So proud was Demosthenes of his preparations that he sometimes even men-
tioned the notes he had made.187 But such forethought was not exceptional. The
sophist Alcidamas mocked those who wrote out their speeches in full, yet did not
recommend speaking offhandedly. Public speakers ought to prepare carefully,
choosing in advance their arguments and overall organization; only the actual
words should be supplied at the time of speaking.188 Even those (few, Demosthe-
nes implies) who came forward on the spur of the moment came because a timely
suggestion had already occurred to them.189
Addressing the assembly was thus typically conceived not as participating in
dialogical deliberation but as offering advice to those who were still deliberating.
And that deliberation was internal. In 346, two assembly meetings were held con-
cerning peace with Philip of Macedon: the first featured speeches, the second only
a series of votes, but the dmos was described as deliberating on both occa-
sions.190 Second person plural imperatives indicating mental activity also com-
monly appeared alongside bouleuesthe.191 Also striking is a speech made by a
Corcyraean envoy to the assembly early in Thucydides where bouleuomai appears
in the singular. If anyone thinks that this course isinexpedient, but fears that if
he yieldshe will be breaking off the trucehe should understandthat he is

182
Dem. Ex. 18. Cf. Ex. 10, 47, 56.
183
Dem. 1.1, 13.3, Ex. 56.
184
Isoc. 13.8, trans. Norlin. Note the use of the first person plural to refer to speakers alone.
185
Aeschin. 3.128, trans. Adams. Cf. Xen. Hell. 1.1.30; Dem. 21.74.
186
Dem. Ex. 33. Cf. Lys. 14.45.
187
Dem. 21.130, 191. Cf. Aeschin. 2.35.
188
Alc. 2.33, trans. Gagarin and Woodruff 1995.
189
Dem. 1.1 = Ex. 3.1.
190
Aeschin. 2.65-8; Dem. 19.13.
191
E.g. skepsasthe, consider, Andoc. 1.144, 3.2, 17, Aeschin. 2.7, 51-2, 153; skopeite, ex-
amine, Aeschin. 2.69, 160. 3.120, 176, Dem. 13.2, 15.26; logizesthe, reflect on, Dem. 8.19,
13.2, 15.17.

24
deliberating (bouleuomenos) upon the interests, not so much of Corcyra, as of
Athens.192 Evidently, the speaker took it for granted that each listener was, at that
very moment, coming to a decision internally.

The collective agent and its rule

As noted, the activity of the assembly recalls guided deliberation, where the deci-
sion-maker takes advice from others. But in one respect it differs significantly
from the examples of guided deliberation seen above. The division of political
labour between, say, Xerxes and Artemisia was absolute: Xerxes decided, Arte-
misia advised. But at the end of his speech, a rhtor returned to his seat and, when
called to vote, raised his hand along with everyone else. He thus took an equal
part in decision-making, even if he played the more rarefied role of advisor during
the consideration stage. What then was the relationship between speaker and au-
dience? Were speakers really not conceived as part of the deliberating dmos?
The cleavage between speaker and audience that opened up when a man came
forward to speak does seem to have persisted beyond the duration of his speech.
Presumably it made some difference how frequently one spoke, but speakers and
non-speakers were conventionally represented as very different groups. Pericles
distinguished those who had besides politics, their private affairs to attend to
from others who though occupied with the pursuits of industry, are still fair judg-
es of public matters.193 Isocrates called orators a tribe (genos) opposed to the
majority (plthos), while Hypereides characterized speakers as snakes distinct
from men, though some at least could be useful.194
One basis for the distinction was the different responsibilities of speakers and
audiences. Orators made proposals, took a broad view and explored best poli-
cy, discerned the trend of events, forecast results and offered warnings when
necessary.195 By contrast, the right course for you, Demosthenes said, is first to
hear the facts, next to decide (bouleusasthai), and finally to carry out your deci-
sion.196 His obligation was to tell you what I have convinced myself is advanta-
geous, his audiences to listen, to judge, and if it is your pleasure, to adopt.197
Most significantly, the audience was responsible for the actions of the polis. Who
sent reinforcements to Byzantium and prevented the entrapment of the Helles-
pont? Demosthenes asked, and answered: You, and when I say you I mean the
polis. He went on: Who advised the polis, moved resolutions, took action? I
did.198
This division of political labour had a legal underpinning. Speakers were not
responsible for the polis actions, but they were accountable for their advice, and
the penalties might be severe. Long after their interventions, they could be indict-
ed for having made an illegal proposal, for deceiving the dmos, or for having

192
Thuc. 1.36, trans. Smith.
193
Thuc. 2.40.2, trans. Crawley. Cf. 3.37-8, 43, 6.39; Loraux 2006, pp. 229-33.
194
Hyp. frag 19.5, Isoc. 8.129. Cf. Hyp. 4, Isoc. 8.26; Dem. 21.189-90; Aeschin. 2.74; Lyc
1.31; Din. 3.19. Cf. Ober 1989, pp. 104-18.
195
Din. 1.35; Dem. 16.1, 18.246.
196
Dem. 19.34, trans. Vince.
197
Dem. Ex. 33, trans. DeWitt.
198
Dem. 18.88, trans. Vince. Cf. Ex. 4; Aeschin. 2.160.

25
spoken when prior immoral behaviour prohibited it.199 By contrast, voting carried
no risk at all. As Diodotus said, we are accountable for our advice to you, but
you who listen are accountable to nobody.200
This system has struck some as inconsistent, even absurd. Why, especially,
should voters blame their advisors for advocating actions that they had approved?
As Thucydides observed, when the Athenians heard of the disaster in Sicily in
413, they were angry at the orators who had taken part in promoting the expedi-
tionas though they had not voted for it themselves.201 Yet Demosthenes al-
lowed that his liability was a fair exchange for political influence. No man
iscompelled to handle politics, he argued. When a man puts himself for-
ward[you] give him appointments and entrust him with public business. If he
succeeds, he will be honoured, and so far gain an advantage over ordinary people;
but if he fails, shall he put forward excuses and apologies? That would be un-
fair and poor consolation to those he had ruined.202
Another way to put this point is to say that speakers chose to play an individu-
al role in a political system defined by the rule of the collective, i.e. the dmos. In
doing so they knowingly cast themselves outside that collective.203 And in disci-
plining its advisors while maintaining its own inviolability the Athenian dmos
was only enjoying one of the prerogatives of kratos, superior power or rule.
Xerxes, Darius and Philip treated their advisers with exactly the same freedom.204
The direct evidence identifying the dmos with the audience in the assembly is
strong. It is implied in the verb dmgore, address the dmos, encountered
above. Speakers also spoke en t dm, in or among the dmos, or pros ton
dmon, before the dmos, and were called ho tou dmou symboulos, counsellor
of the dmos.205 Dmos was the direct object of speech and often equated with
you. Pros ton dmon was equivalent to before you, and the deliberating
dmos to you deliberating.206
Brief consideration of the term dmos explains why speakers should have
been regarded as distinct. Dmos is a collective noun that takes only singular
verbs. By definition, it performed only collective action: not only deliberating but
listening, shouting, judging, voting, and acting, all capable of being performed by

199
Hansen 1974, 1975; Aeschin. 1, Dem. Ex. 6.
200
Thuc. 3.43, trans. Smith. Cf. Ps. Xen. Ath. Pol. 2.17; Hyp. 4.4, 9, 27, 31.9; Dem. 19.182,
21.189-90, 23.4, 24.123, 25.41, 26.4-6, Ex. 25; Aeschin. 1.194-5, 2.160; Landauer 2013.
201
Thuc. 8.1, trans. Smith. Dem. 23.97; Ps. Aristot. Ath. Pol. 28.3; Knox 1985.
202
Dem. 19.103-4, trans. Vince. Cf. 23.97; Hyp. 4.9.
203
Finley 1985, p. 62 (quoting Grote); Ober 1989, p. 295.
204
The parallel between the dmos and a monarch was not lost on contemporaries. Aeschines
observed that in dmokratia, through law and vote, the idiots (ordinary citizen) is king (2.33),
while Aristotle described extreme dmokratia as a system in which the dmos becomes a mon-
arch and demagogues arise (Pol. 1292a10).
205
Andoc. 1.11, 75, Lys. 16.20, Isaeus 11.48, Aeschin 1.20, 27-8, 35, 64, 80, 2.17, 25, 47,
Dem. 19.114, 135, 182, 234, 22.59, 61, 23.172, 58.62, Din. 2.12, 16-17, Hyp. 5 col. 25; Hdt. 9.5,
Thuc. 5.27, Aeschin. 1.81, 2.43, 3.220; Aeschin. 1.120.
206
Thuc. 5.28; Plat. Gorg. 502c-e; Aeschin. 1.25, 81-2, 84-5, 2.43, 59-61, 79, 3.251-2; Dem.
23.97; Ps. Aristot. Ath. Pol. 22.7.

26
a crowd.207 The ancient Greeks did not lack a term for people conceived as a
disaggregated multitude: laoi, a plural noun.208 But the dmos, not laoi, ruled in
dmokratia. Still, dmokratia featured several roles that could only be performed
by a single man, such as general, treasurer, envoy, prosecutor, and rhtor.209
Orators were then a structural necessity, as Moses Finley argued long ago.210
Someone had to put forward and advocate proposals, go on missions abroad, and
generally play a larger political role than ordinary voters. One Greek term for this
role was dmaggos, literally leader of the people (from dmos and ag,
lead).211 There was nothing inherently pernicious about the Athenian dema-
gogue. But by definition, he was not a member of the dmos; and as a result, he
was not regarded as deciding policy.212
It is true, of course, that coming forward to speak did not entail losing ones
vote. But being one in tens of speakers was so much more conspicuous than being
one in thousands of voters that we should not be surprised if the former seems to
have overshadowed the latter. Moreover, bouleuomai indicated coming to a deci-
sion, but as we have seen, those who came forward to advise the dmos had al-
ready decided how they would vote, and their speeches announced their position.
This is not necessarily the case in modern dialogical-deliberative contexts: there,
everyone is assumed to be open to changing their minds down to the last mi-
nute.213 But Athenian political speech was partisan and monological, not in the
sense that there were no interruptionsas we have observed, shouting was com-
monbut in the sense that each speaker offered a single connected argument ei-
ther for or against a proposed course of action. Symbouleutic (that is, advisory)
speeches emphatically did not resemble philosophical dialectic.
For an orator to vote against the position he had just recommended would thus
have been an extraordinary volte-face, as incredible as a Republican candidate for
office publicly endorsing his Democratic opponent. It follows that when orators
voted, they were effectively repeating themselves. They remained cut off from the
crowd even in the decision-making moment, because they had already revealed
their hands. If an audience-members voice was his vote, speakers votes were
their voices, and they cast them before everyone else.
We may conclude that, as Bryan Garsten has suggested, the unity of the
dmos did indeed rest on common deliberationjust not the dialogical kind.214
Dmos implied those who came to a decision as a collective. And although this
process involved many collective acts, the most important was voting, for three
reasons. One, it was above all the act of simultaneously raising hands (assessed by

207
Din. 2.16, Dem. 2.31, 19.8, 34, 234; Aeschin. 1.85; Aristot. Rh. 1.4, Thuc. 6.39; Andoc.
1.77, Aeschin. 2.60, 160, Dem. 8.19; Dem. 19.34, 18.88, Hyp. 4.9.
208
See further Cammack forthcoming.
209
Cf. Aristot. Pol. 1281b25-30.
210
Finley 1985, pp. 38-75 (first published 1962).
211
E.g. Din. 1.40, 72, 98. In addition to Finley (cited above), see Connor 1972, pp. 108-10
(see also 110-15 on prostats tou dmou, one who stands before the dmos); Ober 1989, p. 106;
Rhodes 2016.
212
As emphasized by Finley 1985, pp. 24-25; Hansen 1989, p. 3.
213
Cohen 1989; Fishkin 2009, p. 30; Manin 1987; Dryzek 2000.
214
Garsten 2013.

27
estimation, significantly, not counted) that rendered the collective identity of the
decision-maker visible.215 Two, every other collective act performed prior to the
vote depended on it: there would have been no need to attend, listen, think and so
on had the dmos not been about to make a decision. And three, it was the vote of
the dmos that decided the actions of the polis. In this respect, it seems a mistake
to label those who listened and voted in the assembly passive participants. The
agency of the polis was lodged in their deliberations, even if they did not speak.

Conclusion

Classical Athenian democracy was complex and care is needed to distinguish be-
tween the different tasks involved. The foregoing analysis clarifies the division of
political labour in an area of special interest to political theorists. Both judges and
assembly audiences deliberated (bouleuomai), i.e. came to decisions about actions
within their power. This deliberation was internal, apart from shouting and brief
exchanges, and communicated through voting. Assembly speakers addressed the
dmos (dmgore) and advised (symbouleu). The council prepared material
(bouleu, probouleu) for decision-making by the assembly and courts. This ac-
tivity may have been dialogical, but it was not deliberative on the ancient Greek
conception because it did not involve deciding the matter under consideration.
Conversations about politics did take place outside political institutions, in the
agora, shops, stoa, and schools, for example.216 And they surely influenced later
decisions, just as similar interactions do today. But they were called not bouleusis,
deliberation, but dialogos, syllogos or logos, conversation, discussion or
speech, again because they did not involve decision-making. And they do not
seem to have been regarded as part of a wider system of distributed delibera-
tion.217 Whether a decision was about to be made was assumed to make a signifi-
cant difference. As Isocrates wrote, mockingly: What we condemn before we
enter the assembly, we vote for when we are in sessionand a little later, when
we go home, we disapprove of the resolutions we made there.218 Demosthenes,
too, remarked on the gulf between informal conversation and formal meetings.
Before we come upany one of youis prepared to say readily by what means
the present state of affairs may be improved; and then again, the minute you leave
the assembly each man is just as ready to say what we ought to do. But when we
meet togetheryou hear anything but this from certain speakers.219
The examples we have of dialogical deliberation in political settings in ancient
Greece are all oligarchical. Two have already been discussed: the cases of the
Persian noblemen in Herodotus and the council of kings in Platos Critias. Addi-
tionally, Xenophon used bouleuomai in connection with Athens Thirty Tyrants,
who arrived at decisions through discussion in council.220 Platos Laws showcases
215
Schwartzberg 2010.
216
Antiph. 6.39, 40; Isoc. 18.9; Dem. 19.122, 21.4; Aeschin 1.94, 3.1; Din. 1.32. Connor
1972, 64-6; Finley 1985, pp. 22, 54; Vlassopoulos 2007.
217
Mansbridge et al. 2012. Cf. Owen and Smith 2015.
218
Isoc. 8.52.
219
Dem. Ex. 14, trans. Vince.
220
Xen. Hell. II.3.13, cf. e.g. 3.27; Lys 12.25, 50.

28
a form of participatory oligarchy: young men were to offer advice (symboulia)
while the wise old men would decide (bouleuesthai). Aristotle described a
broader form: all could take part in advising (symbouls) but only elected magis-
trates decide (bouleuesthai), possibly after discussion.221 These examples make
sense: oligarchy means rule by small numbers, and small numbers may deliberate
dialogically. Yet they are hardly encouraging for democrats.
What can political theorists take from this evidence? For one thing, it suggests
that the claim that dialogue is an essential feature of democracy is relatively nov-
el, perhaps no older than the nineteenth century. Consider, for example, John Stu-
art Mills view that the proper function of a deliberating body was talking rather
than doing.222 Why should this have seemed plausible, after the long-standing
association of deliberation with thought and action? The expansion of the fran-
chise may well be relevant. The compatibility of representative and deliberative
democracy is now accepted; I would suggest that they are logical bedfellows.223 It
cannot be a coincidence that the association between deliberation and decision-
making became weaker around the same time that large numbers of ordinary citi-
zens were being incorporated into the political process, albeit not in order to de-
cide law and policy themselves, but rather to decide who would do so. Yet these
citizens could still talk, not only to one another but also to their representatives,
including via the media and public protests, and such communicative action grad-
ually came to seem a more important channel of popular political agency than
voting. The modern interest in deliberative democracy, on this view, is just what
arises when policy-making by ordinary voters has been ruled out, but some sort of
participation beyond voting in elections still seems desirable.
The beginnings of this view appeared in Mills writing on Athens. The Athe-
nian political system had thecharacteristic, far more practically important than
even the political franchise; it was a government of boundless publicity and free-
dom of speech.224 To be sure, Mills conception of dialogical deliberation within
representative government differed from later models: the conversations of the
nations elected representatives would provide the fodder for those who drafted
the laws. These representatives would retain decision-making power in the form
of an eventual vote on legislation, but importantly, in Mills presentation, this de-
cision-making power supplemented their discursive functionit was not intrinsic
to it.225 Contrast this with Burkes exclamation, What sort of reason is that, in
which the determination precedes the discussion; in which one sett [sic] of men
deliberate, and another decide?226 To Burke, the decoupling of deliberation
from decision-making was self-evidently absurd. Yet Mill helped to advance it
and strengthened the association of deliberation with dialogue.
The Athenians reliance on guided deliberation prompts another question.
What is gained by speaking about something oneself as opposed to listening to
221
Plat. Laws 965a. Aristot. Pol. 1298b28. Cf. Plat. Laws 674b, Isoc 3.19.
222
Mill 1991 [1861], pp. 432-3. For discussion, see Urbinati 2002.
223
Gutmann and Thompson 1996, p. 131; Mansbridge and Parkinson eds. 2012; Urbinati
2000, 2002.
224
Mill 1978 [1856], p. 324.
225
Mill 1991 [1861], pp. 424, 430.
226
Burke 1999 [1774], p. 11.

29
others and thinking about it? Presumably it does make a difference to ones en-
gagement with an issue to enter a meeting expecting to participate publicly rather
than internally or as one of a crowd; and certainly, whoever does not speak de-
prives everyone else of the benefit of their thoughts. On the other hand, since
there is seldom enough time for everyone to speak, some compromise on this
score seems inevitable; moreover, privileging speech actsespecially of a nar-
rowly defined kindputs many people at a disadvantage,227 and recent evidence
suggests that more is gained from being exposed to information about an issue
than from discussing it.228
Most interesting, even those, such as James Fishkin, who are dedicated to ex-
panding participation in discussion often write as though the main goal is thought.
Our subject is how to achieve deliberative democracy: how to include everyone
under conditions where they are effectively motivated to really think about the
issues.229 Does such motivation arise only when group discussion takes place?
Oras the Athenian case impliesmight participating in a decision prove an ad-
equate incentive? After all, huge numbers of people attended the assembly and
courts. Only the council struggled to attract members, and the council, of course,
was the only major institution that did not decide political issues itself.230
Still, as noted above, the main reason to insist on dialogical deliberation has
been to support the legitimacy of democracy as a regime. If many views are aired,
justified and interrogated publicly, it is said, the chances of arriving at mutual un-
derstanding and thus reasonable outcomes increase. On this logic, the decisions of
democracies are legitimate to the extent that they are based on dialogue. The deci-
sion-making process determines the legitimacy of the result.
In the ancient Greek evidence, this kind of argument is largely conspicuous by
its absence. Only Plato, as far as one can tell, sought to connect the legitimacy of
regimes to the processes through which decisions were made; even Aristotle did
not think it illegitimate for a dmos to hold kratos and to decide whatever and
however it pleased, though he still believed that democracy was defective and
could lead to unjust actions.231 On the usual ancient Greek conceptualization, it
was the decision-maker that determined the deliberative process, not the delibera-
tive process that determined who decided. In monarchies, the norm was guided
deliberation; in oligarchies, dialogical; in democracies, guided deliberation again.
No attempt was made, except by Plato (most obviously in the Gorgias), to evalu-
ate these models against each other with respect to any external benchmark.
The modern interest in the legitimizing effects of public reason-giving may
seem an improvement over ancient Greek norms. It has certainly helped to ground
a specifically liberal conception of democracy. But political theorists focus on
dialogue has also drawn attention away from what, to an ancient Greek democrat,
was the entire point of democratic deliberation: the making of political decisions
by the dmos. As Cristina Lafont has suggested, small-scale, politically effective

227
Sanders 1997; I.M. Young 1996.
228
Goodin 2008, pp. 38-63; cf. Landemore 2012, pp. 130-6.
229
Fishkin 2009, p. 1 (italicisation of think mine). Cf. pp. 2, 27, 30, 40 et passim.
230
Rhodes 1972, pp. 3-4; Hansen 1999, 248-9.
231
On Plato, see Cammack 2015. For Aristotle, see Pol. 1279a-b, 1281a.

30
dialogical deliberation has an oligarchical tendency.232 The ancient Greek evi-
dence supports that view. Indeed, the form of participatory oligarchy mentioned
by Aristotle, in which all may take part in advising but elected magistrates decide,
is not so distant from some versions of deliberative democracyor from repre-
sentative democracy more generally.
Yet if on this account Athenian democracy looks less useful for proponents of
dialogical deliberation, its value for political theorists may be increased. At least
in respect of public speech, classical Athenian and modern democratic politics
would seem more similar than is often recognized. In both cases, an overwhelm-
ing majority of citizens engage in politics through voting and informal conversa-
tion, while a small number take leading political roles. The crucial differences
concern not the role of dialogue but the fact that in Athens, large groups of ordi-
nary voters decided all political issues; that the formal bar to speaking publicly
was low; and that the risks associated with speaking publicly were high, thanks to
the dmoss use of the courts to discipline politicians. This is the opposite of the
case today, when a high barrier to entry as a politician (often financial, but also
social and cultural) is combined with a low risk of being held personally account-
able for ones actions.
In an important argument, Josiah Ober suggested that direct democratic deci-
sion-making in classical Athens was made possible by the mediating and integra-
tive power of communication between citizensespecially between ordinary and
elite citizens.233 It was the ideological hegemony of the mass over the elite, re-
vealed in the willingness of orators to bow, rhetorically, to mass audiences, that in
Obers opinion ensured the preservation of the democratic political system.234
Obers conviction that in Athens, the mass really did rule is, I think, entirely mer-
ited. Nonetheless, the only reason orators needed to appeal rhetorically to mass
audiences was because decision-making power was in those audiences hands. In
that respect, Obers argument illustrates a more general tendency to minimize the
significance of formal institutional control as opposed to communicative action
among scholars in recent decades. What secured the rule of the Athenian dmos
over its political elite was its jealous preservation of its right to deliberate, i.e.
make its own decisions, in both the assembly and courts. This entailed using a
type of deliberation appropriate to the size of those institutions: internal delibera-
tion, guided by a small number of advisory speeches, concluding in a vote. The
only Athenian whom we know lamented the resulting lack of dialogos within the
political system was Plato, and he is hardly a model for democrats today.

References

Abramson, Jeffrey. The jury and democratic theory. Journal of Political Philosophy, 1,
45-68.
Allen, Danielle. 2000. The World of Prometheus. Princeton, N.J.: Princeton University
Press.

232
Lafont 2015.
233
Ober 1989, p. 35.
234
Ober 1989, pp. 156, 191, 307, 328-38.

31
Arendt, Hannah. 1998 [1958]. The Human Condition. Chicago: Chicago University
Press.
Bchtiger, Andr, Simon Niemayer, Michael Neblo, Marco Steenbergen and Jrg Steiner.
2010. Disentangling diversity in deliberative democracy. Journal of Political Philos-
ophy, 18, 32-63.
Balot, Ryan. 2004. Greek Political Thought. Oxford: Blackwell.
Benhabib, Seyla. 1996. Toward a deliberative model of democratic legitimacy. Pp. 67-94
in Benhabib ed. 1996.
Benhabib, Seyla ed. 1996. Democracy and Difference: Contesting the Boundaries of the
Political. Princeton, N.J.: Princeton University Press.
Bers, Victor. 1985. Dikastic thorubos. Pp. 1-15 in Paul Cartledge and F.D. Harvey (eds),
Crux: Essays in Greek History. London: Duckworth.
Bickford, Susan. 1996. The Dissonance of Democracy. Ithaca, N.Y.: Cornell University
Press.
Boegehold, Alan. 1995. The Lawcourts at Athens. Princeton: Princeton University Press.
Bohman, James. 1996. Public Deliberation. Cambridge MA: MIT Press.
Burke, Edmund. 1999 [1774]. Speech to the electors of Bristol. Pp. 3-14 in Francis Ca-
navan (ed), Select Works of Edmund Burke vol. 4: Miscellaneous Writings. Indianap-
olis, IN: Liberty Fund.
Cammack, Daniela. 2013a. Aristotle on the virtue of the multitude. Political Theory 41,
175-202.
Cammack, Daniela. 2013b. Aristotles denial of deliberation about ends. Polis, 30, 228-
50.
Cammack, Daniela. 2015. Plato and Athenian justice. History of Political Thought, 36,
611-42.
Cammack, Daniela. Forthcoming. The Dmos in Dmokratia. Available on SSRN.
Chambers, Simone. 2009. Rhetoric and the public sphere: has deliberative democracy
abandoned mass democracy? Political Theory, 37, 323-50.
Chambers, Simone. 2012. Deliberation and mass democracy. Pp. 52-71 in John Parkinson
and Jane Mansbridge (eds) 2012.
Cohen, Joshua. 1989. Deliberation and democratic legitimacy. Pp. 67-92 in Alan Hamlin
and Philip Pettit (eds), The Good Polity. Oxford: Blackwell.
Connor, W. Robert. 1972. The New Politicians of Fifth-Century Athens. Princeton, N.J.:
Princeton University Press.
Dahl, Robert. 1989. Democracy and its Critics. New Haven, CT: Yale University Press.
Deslauriers, Marguerite and Pierre Destre eds. 2013. The Cambridge Companion to Ar-
istotles Politics. Cambridge: Cambridge University Press.
Dryzek, John. 2000. Deliberative Democracy and Beyond. Oxford: Oxford University
Press.
Dunn, John. 2005. Democracy: A History. New York: Atlantic Monthly Press.
Elster, Jon. 1998. Introduction. Pp. 1-18 in Jon Elster (ed), Deliberative Democracy.
Cambridge: Cambridge University Press.
Finley, Moses. 1985. Democracy Ancient and Modern. New Brunswick: Rutgers Univer-
sity Press.
Fishkin, James. 2009. When the People Speak. Oxford: Oxford University Press.
Fishkin, James and Peter Laslett eds. 2003. Debating Deliberative Democracy. Oxford:
Blackwell.
Forsn, Bjrn and G. Stanton eds. 1996. The Pnyx in the History of Athens. Athens,
Greece: Foundation of the Finnish Institute at Athens.
Gagarin, Michael and Paul Woodruff. 1995. Early Greek Political Thought from Homer
to the Sophists. Cambridge: Cambridge University Press.

32
Garsten, Bryan. 2006. Saving Persuasion. Cambridge, MA: Harvard University Press.
Garsten, Bryan. 2013. Deliberating and acting together. Pp. 324-49 in Deslauriers and
Destre eds. 2013.
Grote, George. 1846-56. History of Greece. 10 vols. London: John Murray.
Goodin, Robert. 2003. Reflective Democracy. Oxford: Oxford University Press.
Goodin, Robert. 2008. Innovating Democracy. Oxford: Oxford University Press.
Gutmann, Amy and Dennis Thompson. 1996. Democracy and Disagreement. Cambridge,
MA: Harvard University Press.
Gutmann, Amy and Dennis Thompson. 2004. Why Deliberative Democracy? Princeton,
N.J.: Princeton University Press.
Habermas, Jrgen. 1985. The Theory of Communicative Action, 2 vols., trans. Thomas
MacCarthy. Cambridge MA: MIT Press.
Habermas, Jrgen. 1989. The Structural Transformation of the Public Sphere, trans.
Thomas Burger. Cambridge, MA: MIT Press.
Habermas, Jrgen. 1996. Between Facts and Norms, trans. William Rehg. Cambridge
MA: MIT Press.
Hansen, Mogens Herman. 1983. The Athenian Ecclesia. Copenhagen: Museum Tuscu-
lanum Press.
Hansen, Mogens Herman. 1989. The Athenian Ecclesia II. Copenhagen: Museum Tuscu-
laneum Press.
Hansen, Mogens Herman. 1999. Athenian Democracy in the Age of Demosthenes. Nor-
man, OK: University of Oklahoma Press.
Hansen, Mogens Herman. 2010. The concepts of demos, ekklesia, and dikasterion in
classical Athens. Greek, Roman and Byzantine Studies, 50, 499-536.
Hardie, W.F.R. 1980. Aristotles Ethical Theory. Oxford: Clarendon Press.
Held, David. 2006. Models of Democracy. Palo Alto: Stanford University Press.
Irwin, Terence. 1975. Aristotle on Reason, Desire and Virtue. Journal of Philosophy, 72,
567-78.
Johnstone, Christopher Lyle. 1996. Greek oratorical settings and the problem of the Pnyx.
Pp. 97-127 in Christopher Lyle Johnson (ed), Theory, Text, Context: Issues in Greek
Rhetoric and Oratory. Albany, N.Y.: State University of New York Press.
Kallet-Marx, Lisa. 1994. Money talks: rhetor, demos, and the resources of the Athenian
empire. Pp. 227-52 in S. Hornblower and Robin Osborne (eds), Ritual, Finance, Poli-
tics: Athenian Democratic Accounts presented to David Lewis. Oxford: Oxford Uni-
versity Press.
Knox, Ronald. 1985. So mischievous a beaste? The Athenian demos and its treatment
of its politicians. Greece & Rome, 32, 132-61.
Landauer, Matthew. 2012. Parrhesia and the demos tyrannos. History of Political
Thought, 33, 185-208.
Landauer, Matthew. 2013. The idiots and the tyrant: two faces of unaccountability in
democratic Athens. Political Theory, 42, 139-66.
Landemore, Hlne. 2012. Democratic Reason: Politics, Collective Intelligence, and the
Rule of the Many. Princeton, N.J.: Princeton University Press.
Lafont, Cristina. 2015. Deliberation, participation, and democratic legitimacy: should
deliberative mini-publics shape public policy? Journal of Political Philosophy, 23,
40-63.
Lane, Melissa. 2013. Aristotle on claims to rule. Pp. 247-74 in Deslauriers and Destre
eds. 2013.
Lewis, J.D. 1971. Isgoria at Athens: when did it begin? Historia, 20, 129-40.
Liddell, Henry, Robert Scott and Henry Stuart Jones. 1996. Greek-English Lexicon. Ox-
ford: Oxford University Press.

33
Loraux, Nicole. 2006. The Invention of Athens, trans. A. Sheridan. New York: Zone
Books.
Manin, Bernard. 1987. On legitimacy and political deliberation. Political Theory, 15,
338-68.
Manin, Bernard. 1997. The Principles of Representative Government. Cambridge: Cam-
bridge University Press.
Manin, Bernard. 2011. Comment promouvoir la dlibration dmocratique? Raisons
Politiques, 43, 83-113.
Mansbridge, Jane, James Bohman, Simone Chambers, Thomas Christiano, Archon Fung,
John Parkinson, Dennis Thompson and Mark Warren. 2012. A systemic approach to
deliberative democracy. Pp. 1-26 in John Parkinson and Jane Mansbridge (eds) 2012.
Mansbridge, Jane and John Parkinson. 2012. Deliberative Systems: Deliberative democ-
racy at the large scale (Cambridge: Cambridge University Press).
Marr, J.L. and Peter J. Rhodes. 2008. The Old Oligarch: The Constitution of the Athe-
nians. Oxford: Aris & Phillips/Oxbow.
Mill, John Stuart. 1991 [1861]. Considerations on representative government. Pp. 203-
467 in John Gray (ed), On Liberty and Other Essays. Oxford: Worlds Classics.
Mill, John Stuart. 1978 [1856]. Grotes History of Greece II. Pp. 307-39 in The Collected
Works of John Stuart Mill, Vol. XI: Essays on Philosophy and the Classics. Toronto:
University of Toronto Press.
Miller, Jeff. 2001. Democratic characterizations of liberty. History of Political Thought,
22, 400-17.
Monoson, Sara. 2000. Platos Democratic Entanglements: Athenian Politics and the
Practie of Philosoophy. Princeton, N.J.: Princeton University Press.
Mulgan, Richard. 1999. Aristotle, ethical diversity and political argument. Journal of Po-
litical Philosophy 7, 191-207.
Ober, Josiah. 1989. Mass and Elite in Democratic Athens. Princeton: Princeton Universi-
ty Press.
Ober, Josiah. 2008. Democracy and Knowledge. Princeton: Princeton University Press.
Ober, Josiah. 2013. Democracys wisdom: an Aristotelian middle way for collective
judgment. American Political Science Review, 107, 104-22.
Osborne, Robin. 1985. Demos: The Discovery of Classical Attika. Cambridge: Cam-
bridge University Press.
Owen, David and Graham Smith. 2015. Deliberation, democracy and the systemic turn.
Journal of Political Philosophy, 23, 213-34.
Pirart, Marcel. 1974. A propos de llection des stratges athniens. Bulletin de corre-
spondance hellnique, 98, 125-46.
Radin, Max. 1927. Freedom of speech in ancient Athens. American Journal of Philology,
48, 215-30.
Rawls, John. 1999. A Theory of Justice. Cambridge, MA: Belknap Press.
Remer, Gary. 1999. Political oratory and conversation. Political Theory, 27, 39-64.
Remer, Gary. 2000. Two models of deliberation. Journal of Political Philosophy, 8, 68-
90.
Rhodes, Peter J. 1972. The Athenian Boule. Oxford: Clarendon Press.
Rhodes, Peter J. 1981. Notes on voting in Athens. Greek, Roman and Byzantine Studies,
22, 129-32.
Rhodes, Peter J. 1995. The acephalous polis? Historia, 44, 153-67.
Rhodes, Peter J. 2016. Demagogues and Demos in Athens. Polis, 33, 243-64.
Roisman, Joseph. 2004. Speaker-audience interaction in Athens: a power struggle. Pp.
261-78 in Sluiter and Rosen (eds) 2004.

34
Ruz, Franoise. 1997. Dlibration et Pouvoir dans la Cit Grecque. Paris: Publ. de la
Sorbonne.
Sanders, Lynn. 1997. Against deliberation. Political Theory, 25, 347-76.
Saxonhouse, Arlene. 2006. Free Speech and Democracy in Ancient Athens. Cambridge:
Cambridge University Press.
Schwartzberg, Melissa. 2010. Shouts, murmurs, and votes. Journal of Political Philoso-
phy, 18, 448-68.
Sintomer, Yves. 2010. Random selection, republican self-government, and deliberative
democracy. Constellations, 17, 472-
Smyth, Herbert Weir. 1956. Greek Grammar. Cambridge MA: Harvard University Press.
Sluiter, Ineke and Ralph Rosen eds. 2004. Free Speech in Classical Antiquity. Leiden:
Brill.
Stevens, Edward Boucher. 1933. The topics of counsel and deliberation in prephilosophic
Greek literature. Classical Philology, 28, 104-20.
Tacon, Judith. 2001. Ekklesiastic thorubos. Greece & Rome, 48, 173-92.
Thompson, Dennis. 2008. Deliberative democratic theory and empirical political science.
Annual Review of Political Science, 11, 497-520.
Urbinati, Nadia. 2000. Representation as advocacy: a study of democratic deliberation.
Political Theory, 28, 758-86.
Urbinati, Nadia. 2002. Mill on Democracy: From the Athenian Polis to Representative
Government. Chicago: Chicago University Press.
Urfalino, Philippe. 2005. La dlibration nest pas une conversation. Ngociations, 4, 99-
114.
Urfalino, Philippe. 2010. Deciding as bringing deliberation to a close. Social Science In-
formation, 49, 110-40.
Villacque, Nomie. 2012. Chahut et dlibration. Participations, 3, 49-69.
Vlassopoulos, Kostas. 2007. Free spaces: identity, experience and democracy in Classical
Athens. Classical Quarterly, 57, 33-52.
Waldron, Jeremy. 1999. The Dignity of Legislation. Cambridge: Cambridge University
Press.
Wallace, Robert. 2004. The power to speakand not to listen. Pp. 221-32 in Sluiter and
Rosen (eds) 2004.
Whitehead, David. 1986. The Demes of Attica. Princeton, N.J.: Princeton University
Press.
Yack, Bernard. 2006. Rhetoric and public reasoning: an Aristotelian understanding of
political deliberation. Political Theory, 34, 417-38.
Yunis, Harvey. Taming Democracy. Ithaca: Cornell University Press, 1996.
Young, Iris M. 1996. Communication and the other: beyond deliberative democracy. Pp.
120-35 in Benhabib ed. 1996.

35